The office in charge of foreign investment security review (NDRC) has decided to block the foreign acquisition of the Manus project and require the parties to unwind the deal.

Just one sentence—but the meaning is clear: after months of speculation and review, the rumored Manus–Meta deal is officially over.

The decision was made several days before US President Donald Trump’s potential visit to China,and it was reported that China and US is negotiating in mutual investment. It’s not clear whether this case will be on the table of the “investment board”, which was confirmed and promised to be established during the Paris talks between the two countries.

Going back to early March 2025, Manus first took off through an invite-only beta. At one point, invite codes were even being resold on secondary markets. As the hype grew, so did the questions.

The core issue was pretty straightforward:was this a genuine technical breakthrough, or mainly a clever integration of existing models and tools?

In the tech community and media, the debate focused on a few points—how much Manus relied on third-party model APIs, and whether its so-called “general agent” capabilities came from orchestrating tools, or from real algorithmic innovation. But without full technical disclosure or a shared evaluation framework, the debate never really got settled.

In April 2025, Bloomberg reported that Manus’s parent company, Butterfly Effect, raised about $75 million at a roughly $500 million valuation, led by Benchmark, with plans to expand overseas. By June, Chinese media reported that the company was restructuring globally—moving its headquarters to Singapore and scaling down its China team.

Reactions were mixed. Supporters saw it as a rational move to improve global operations. Critics linked it to geopolitical risks, investor pressure, and potential regulatory concerns.

What really shaped the outcome came at the end of 2025.

On December 29, Manus announced on its blog that it would be joining Meta. It outlined some high-level plans—continuing the product, accelerating iteration, and bringing part of the team into Meta. But one key detail—the deal size—was never clearly disclosed. Media estimates ranged from $2–3 billion to higher, but there was no official number.

That’s when regulators formally stepped in.

In January 2026, the Ministry of Commerce of the People’s Republic of China signaled that it was reviewing the deal. Shortly after, at a regular press briefing, it said it would work with other agencies to assess the transaction’s compliance with export controls, technology transfer rules, and outbound investment regulations.

Looking back, that statement carried a lot of weight. It didn’t frame the issue under just one regulatory regime—it put the deal into a broader, cross-cutting compliance framework. That meant the review could touch on everything: technology, deal structure, funding, control, and procedural issues.

By late March, multiple international media outlets reported that some Manus executives had been restricted from leaving China, reinforcing the sense that the review had become serious.

Interestingly, and contrary to what many policy watchers and lawyers expected, the tool ultimately used was not export controls, but foreign investment security review.

This mechanism is sometimes described as China’s version of CFIUS. Many people have heard of it, but few can clearly explain how it works. That’s partly because it’s not a system with highly detailed rules and transparent case law.

At its core, it’s a cross-agency review mechanism led by the NDRC and the Ministry of Commerce. You can think of it as an interagency consultation system—once a transaction touches on national security concerns, it gets pulled into this process.

What triggers it? Broadly two things.

First, sensitive sectors—traditionally defense, energy, and critical infrastructure. But in recent years, the scope has clearly expanded to include data, AI, and digital platforms.

Second, control. It’s not just about majority ownership. If an investor can materially influence decisions—through equity, contracts, board seats, or other means—that can be enough to trigger a review.

What’s more interesting is how the system actually works in practice.

On paper, it’s a structured process: companies file, regulators conduct an initial review, and if needed, move to deeper assessment or even a “special review.” But in reality, a lot happens before anything is formally filed.

Companies often engage regulators early, even at the deal design stage, to get a sense of the likely outcome. If the risk looks too high, the structure gets changed—or the deal never happens. Cases that are formally blocked are relatively rare in public view, but that doesn’t mean the system isn’t powerful. Its real impact often shows up as self-restraint at the front end.

Outcomes can vary: approval, conditional approval (for example, data localization or operational separation), or outright prohibition with a requirement to unwind. And if a deal wasn’t filed in the first place, it can still be investigated after the fact.

If you compare this with the U.S. CFIUS system, the contrast is quite clear. The U.S. approach is more formalised and institutionalised, with clearer rules and precedents. China’s system is more flexible, less transparent, but also more adaptable—and more closely tied to broader industrial and national security priorities.

One more development is worth noting.

On April 24, Bloomberg reported that China is considering restricting top tech companies—including leading AI startups—from accepting U.S. investment without government approval, to prevent sensitive technologies tied to national security from being accessed by foreign investors.

Looking ahead, this review mechanism is only going to become more important. It used to focus mainly on traditional sectors, but it’s now clearly extending into data, AI, and platform economies—the new critical infrastructure of the digital age.

And even though technology export control rules weren’t ultimately used in the Manus case, that doesn’t mean they’re irrelevant. If anything, the “butterfly effect” of this case could reshape how Chinese AI companies think about globalization—from capital and technology to talent—and raise a whole new set of regulatory questions going forward.

Against the backdrop of escalating U.S.–China trade tensions, the global impact of “reciprocal tariffs,” and the Israel–Iran–U.S. conflict, the Regulations represent a major institutional upgrade of China’s counter-sanctions legal toolkit.

Comprising 20 articles, they establish an integrated framework of identification, blocking, and countermeasures, and for the first time formally codify a “Malicious Entity List” and “prohibition orders” at the level of administrative regulation, signaling a shift in China’s approach to foreign-related rule-of-law from passive defense to a more offensive-defensive posture.

The operational logic of the Regulations can be understood as a closed loop moving from identification, to blocking, to countermeasures, and then to implementation and remedies.

At the core is a centrally coordinated “working mechanism” under the State Council, through which different agencies share information, make decisions, and coordinate enforcement, effectively consolidating previously fragmented capabilities into a unified system.

The first step is identification. Led by the State Council’s rule-of-law authorities, foreign measures are assessed based on whether they violate international law, lack a reasonable jurisdictional nexus, or harm China’s national interests or the lawful rights of its citizens and organizations. Once a measure is determined to constitute “improper extraterritorial application,” it may be publicly designated.

Following identification, the key step is blocking. The Regulations stipulate that, in principle, no organization or individual may comply with such foreign measures. This directly reshapes corporate compliance behavior. At the same time, an exemption mechanism allows entities to apply for approval in exceptional circumstances, thereby centralizing ultimate discretion at the state level.

On this basis, China may adopt countermeasures. These include both macro-level tools—such as diplomacy, trade, and investment—and more targeted instruments linked to the Anti-Foreign Sanctions Law of the People’s Republic of China. In particular, the “Malicious Entity List” enables the imposition of measures such as visa restrictions, asset freezes, and limits on transactions and investment.

Implementation is carried out through the working mechanism, with decisions transmitted to relevant enforcement agencies. Regulators are empowered to conduct investigations, inspections, and compliance interviews, and may issue prohibition orders requiring specific entities not to comply with foreign measures, ensuring enforceability at the operational level.

The Regulations also provide for remedies. Affected parties may apply for suspension or removal of countermeasures, while Chinese citizens and companies may bring lawsuits before domestic courts seeking cessation of harm and compensation, thereby incorporating these disputes into China’s judicial system.

Enforcement is backed by penalties. Entities that fail to comply with countermeasures or violate prohibition orders may face restrictions on market access, trade, data flows, and cross-border mobility, as well as fines and potential legal liability.

Compared with earlier rules such as the 2021 blocking measures issued by the Ministry of Commerce, this Regulation introduces the more robust “Malicious Entity List.” Entities placed on the list may face comprehensive restrictions, including entry bans, loss of work or residence status in China, asset freezes, prohibitions on transactions with Chinese entities, limits on trade and investment, restrictions on products entering China, and monetary penalties.

Notably, the Regulation also restricts the provision and cross-border transfer of data and personal information involving listed entities. This effectively cuts off access to China-based data resources, a measure likely to have a deeper and more structural impact on business operations than traditional tools such as visa bans or asset freezes.

The Regulation marks a significant upgrade from the 2021 blocking rules, which were ministerial-level measures. As a State Council administrative regulation, it carries higher legal authority. It also introduces a key doctrinal shift by explicitly asserting China’s right to exercise extraterritorial jurisdiction over conduct with a sufficient nexus to China, moving beyond defensive blocking toward an affirmative jurisdictional claim.

Since 2020, China has gradually built a layered legal framework to respond to foreign extraterritorial measures, including the 2020 rules on the Malicious Entity List, the 2021 blocking rules, the 2021 Anti-Foreign Sanctions Law, and the 2023 Foreign Relations Law. This Regulation fills an important gap at the administrative level and completes a more coherent system.

From a comparative perspective, similar mechanisms exist elsewhere. The European Union, for instance, adopted its Blocking Statute (Regulation No. 2271/96) as early as 1996 to counter U.S. sanctions. However, China’s approach goes further by introducing proactive tools such as the Malicious Entity List, making it more assertive than the EU’s largely defensive model.

If strictly enforced, the Regulation may place companies operating in China—including multinational corporations and their Chinese subsidiaries—in situations of conflicting legal obligations. Where such conflicts cannot be reconciled, firms may be forced to assess which jurisdiction poses greater regulatory risk, significantly affecting compliance strategies.

In practice, companies may face uncertainty in applying the Regulation. While it outlines general criteria for determining “improper” extraterritorial measures, how these standards will be interpreted remains to be clarified through practice. This makes it difficult for firms to predict with certainty whether a foreign law or order will be deemed improper.

The exemption mechanism is designed to alleviate such dilemmas, but its effectiveness will depend on the clarity, transparency, and predictability of approval procedures. At present, businesses are closely watching for further implementing rules or guidance.

This Regulation forms part of China’s broader effort since 2020 to strengthen its foreign-related legal system. Chinese leadership, including President Xi Jinping, has repeatedly emphasized the need to better integrate domestic and foreign-related rule of law, and to use legal tools to respond to geopolitical competition, safeguard national sovereignty, security, and development interests, and shape global governance.

We must advance the rule of law in both domestic and foreign-related matters in a coordinated manner. We should accelerate the strategic layout of foreign-related legal work, promote coordination between domestic governance and international governance, and better safeguard national sovereignty, security, and development interests. We must strengthen law-based thinking and employ legal means to effectively respond to challenges and prevent risks, making comprehensive use of legislation, law enforcement, and judicial measures to carry out our efforts, and resolutely uphold national sovereignty, dignity, and core interests. We should also promote the reform of global governance and advance the building of a community with a shared future for mankind.

——Remarks by Xi Jinping at the First Meeting of the Central Commission for Comprehensively Advancing the Rule of Law in 2020.

Announcement No. 5 (2021) of the Standing Committee of the National People’s Congress has set out a roadmap for the relevant legislative work:

In accordance with the requirements set by the Fourth Session of the 13th National People’s Congress to improve foreign-related legislation, efforts should be coordinated and accelerated to advance such legislation. Focusing on countering sanctions, countering interference, and countering long-arm jurisdiction, China will enrich its legal “toolbox” to address challenges and mitigate risks, and make better use of legal instruments to safeguard national sovereignty, security, and development interests, while protecting the lawful rights and interests of Chinese citizens and organizations.

On July 11, 2024, Yin Baohu, Deputy Secretary-General of the China Law Society and President of Democracy and the Legal System, wrote in the Legal Daily that China should:

Follow the principle of prioritizing urgent needs, further improve laws and regulations on countering sanctions, countering interference, and countering ‘long-arm jurisdiction’; elevate relevant rules and policies into a formal blocking statute at an appropriate time; and promptly issue implementing rules for the Anti-Foreign Sanctions Law, so as to advance the development of China’s legal framework for extraterritorial application.

In parallel, on March 31, 2026, the State Council issued new regulations on safeguarding the security of industrial and supply chains, which also provide for legal countermeasures against discriminatory restrictions imposed by foreign actors. These measures further reinforce China’s emerging system for responding to external economic and legal pressure.

Full translation of the Regulation (unofficial):

Regulations of the People’s Republic of China on Countering Improper Extraterritorial Application of Foreign Laws

Article 1
These Regulations are formulated in accordance with the National Security Law of the People’s Republic of China, the Foreign Relations Law of the People’s Republic of China, the Anti-Foreign Sanctions Law of the People’s Republic of China, and other relevant laws, in order to safeguard national sovereignty, security, and development interests, protect the lawful rights and interests of Chinese citizens and organizations, and uphold the international order based on international law.

Article 2
Work on countering improper extraterritorial application of foreign laws shall implement a holistic approach to national security, coordinate development and security, and balance domestic and international priorities, so as to safeguard the socialist system with Chinese characteristics and promote the building of a more just and equitable global governance system.

Article 3
The People’s Republic of China adheres to an independent foreign policy of peace, opposes hegemonism and power politics, and opposes any country interfering in China’s internal affairs under any pretext or in any manner.

Where a foreign state, in violation of international law and the basic norms governing international relations, adopts improper extraterritorial measures that endanger China’s sovereignty, security, or development interests, or harm the lawful rights and interests of Chinese citizens or organizations, the Chinese government has the right to take corresponding measures.

Article 4
In accordance with the laws of the People’s Republic of China, treaties concluded or acceded to by the People’s Republic of China, or on the basis of reciprocity, the Chinese government has the right to exercise extraterritorial jurisdiction over acts that have an appropriate connection with China, so as to safeguard national sovereignty, security, and development interests and protect the lawful rights and interests of Chinese citizens and organizations.

Where the Chinese government exercises jurisdiction over relevant acts pursuant to the preceding paragraph, and a foreign state also asserts jurisdiction over the same acts, the parties may resolve the matter, on the basis of compliance with international law and the basic norms governing international relations, through treaties, diplomatic channels, or consultations between competent authorities.

Article 5
The State shall establish and improve a working mechanism for responding to improper extraterritorial application of foreign laws (hereinafter referred to as the “working mechanism”), to coordinate such efforts in an integrated manner.

Relevant departments under the State Council shall, in accordance with their respective responsibilities, undertake specific work in responding to improper extraterritorial measures. These departments and other relevant authorities shall strengthen coordination, cooperation, and information sharing in identifying and responding to such measures.

Article 6
The department in charge of rule of law under the State Council, together with other relevant authorities, shall carry out the identification of improper extraterritorial measures, and may conduct investigations and external consultations. Relevant organizations and individuals may submit proposals to initiate such identification work.

In conducting identification, the following factors shall be comprehensively considered:
(1) Whether the measure violates international law and the basic norms governing international relations;
(2) Whether there is an appropriate connection between the conduct subject to extraterritorial jurisdiction and the foreign state concerned;
(3) Whether it endangers China’s sovereignty, security, or development interests, or harms the lawful rights and interests of Chinese citizens or organizations;
(4) Other relevant factors.

Where a measure is identified as constituting improper extraterritorial application of foreign laws, the department in charge of rule of law under the State Council may make an announcement to that effect. No organization or individual may execute or assist in executing such measures.

Where, under special circumstances, a Chinese citizen or organization must execute or assist in executing such measures, it shall apply to the said department, provide relevant facts and reasons, and specify the scope. Upon approval through the working mechanism, such execution or assistance may be permitted within a specified scope.

Article 7
The Chinese government may assess acts by relevant states involving improper extraterritorial measures, determine risk levels, and take countermeasures and restrictive measures in areas such as foreign affairs, entry and exit, trade, investment, international cooperation, and foreign aid, in accordance with law.

Article 8
Relevant departments under the State Council may, pursuant to the working mechanism, include foreign organizations and individuals that promote or participate in the implementation of improper extraterritorial measures in a “Malicious Entity List,” and, in accordance with the Anti-Foreign Sanctions Law and its implementing provisions, decide to adopt one or more of the following countermeasures and restrictive measures, and make them public:
(1) Refusal to issue visas, denial of entry, cancellation of visas, orders to leave within a time limit, repatriation, or deportation;
(2) Cancellation or restriction of qualifications for work, stay, or residence in China;
(3) Seizure, impoundment, or freezing of movable property, immovable property, and other types of assets within China;
(4) Prohibition or restriction on Chinese organizations and individuals from providing data or personal information, or engaging in transactions or cooperation;
(5) Prohibition or restriction on import and export activities related to China;
(6) Prohibition or restriction on investment in China;
(7) Prohibition or restriction on the entry of products, transportation tools, etc.;
(8) Fines;
(9) Other necessary measures.

The above measures may also apply to organizations actually controlled by, or established or operated with the participation of, listed entities or individuals.

Article 9
Organizations or individuals subject to countermeasures and restrictive measures may apply to the relevant State Council departments for suspension, modification, or cancellation of such measures, providing facts and reasons such as corrective actions taken and elimination of consequences.

The relevant departments may assess the implementation and effectiveness of such measures based on actual circumstances.

Based on assessment results or review of applications, the relevant departments may, pursuant to the working mechanism, decide to suspend, modify, or cancel such measures and make the decision public.

Article 10
Where implementation of countermeasures requires action by other State Council departments, the department that made the decision shall notify the relevant departments in accordance with the working mechanism.

Relevant departments shall implement such measures according to their respective responsibilities.

Article 11
Where, under special circumstances, organizations or individuals must engage in otherwise prohibited or restricted activities with entities or individuals subject to countermeasures, they shall apply to the relevant State Council department, provide facts and reasons, and may proceed within an approved scope upon authorization.

Article 12
Relevant State Council departments may take measures such as on-site inspections and reviewing or copying relevant materials against organizations or individuals suspected of executing or assisting improper extraterritorial measures. Relevant parties shall cooperate and shall not refuse or obstruct.

Article 13
Relevant State Council departments may conduct interviews with, and order rectification by, organizations or individuals executing or assisting such measures.

The department in charge of rule of law under the State Council may, pursuant to the working mechanism, issue decisions prohibiting execution of improper extraterritorial measures (hereinafter “prohibition orders”). Relevant parties shall comply.

Article 14
Where any organization or individual executes or assists in executing improper extraterritorial measures, thereby infringing upon the lawful rights and interests of Chinese citizens or organizations, the affected parties may file a lawsuit with a people’s court in accordance with law, seeking cessation of infringement and compensation for losses.

Article 15
Relevant departments of people’s governments at or above the provincial level shall, within their respective responsibilities, provide guidance and services to Chinese citizens and organizations in responding to improper extraterritorial measures.

Article 16
Industry associations and chambers of commerce shall, in accordance with laws, regulations, and their charters, play roles in self-regulation and coordination, guide members to operate in compliance with laws, promptly reflect industry concerns, and provide services in market expansion, rights protection, and dispute resolution related to responding to improper extraterritorial measures.

Article 17
Where any organization or individual refuses to comply with or circumvents countermeasures and restrictive measures, or violates prohibition orders, relevant State Council departments may order rectification, prohibit or restrict participation in government procurement, bidding, import and export of goods and technologies, or international trade in services, restrict cross-border data flows and provision of personal information, restrict entry and exit or stay in China, and impose fines.

Article 18
Where violations of these Regulations constitute a crime, criminal liability shall be pursued in accordance with law.

Article 19
Where matters relating to countering improper extraterritorial measures involve anti-corruption, anti-monopoly, anti-unfair competition, export control, data security, or judicial assistance, and relevant laws and regulations provide otherwise, such provisions shall apply.

Where other provisions exist regarding responses to foreign states that improperly prohibit or restrict normal economic and trade activities between Chinese citizens or organizations and third countries (or regions), such provisions shall apply.

Article 20
These Regulations shall enter into force upon promulgation.

The Ministry of Justice subsequently released a Q&A on the Regulations in the name of “the official in charge of the Ministry of Justice,” and several leading scholars in this field have also published interpretive articles.

According to the Ministry of Justice official, against the backdrop of a volatile international environment and the increasing use of “long-arm jurisdiction” by certain countries to pressure Chinese companies and individuals, it is necessary for China to clearly state, through legislation, its position of rejecting improper foreign extraterritorial jurisdiction and to respond in a systematic, law-based manner. This is not only essential for safeguarding national sovereignty, security, and development interests, but also for protecting the lawful rights and interests of enterprises and citizens and enhancing China’s ability to manage external risks. At its core, it represents an institutionalized response to external legal and geopolitical pressures.

The official further explained that the Regulations adopt a problem-oriented approach and establish a comprehensive framework covering identification, blocking, countermeasures, implementation, and safeguards. On the one hand, they restrict domestic entities from complying with improper foreign laws through identification and blocking measures; on the other hand, they introduce countermeasures—such as state-level tools and the “Malicious Entity List”—while also incorporating supporting mechanisms including prohibition orders, judicial remedies, government assistance, and industry coordination. Together, these elements form an integrated system that both enables external countermeasures and provides internal regulation and support.

Head of the Ministry of Justice Answers Questions from the Press on the Regulations of the People’s Republic of China on Countering Improper Extraterritorial Application of Foreign Laws

Q: Could you briefly introduce the background to the formulation of the Regulations?
A: At present, changes of the world, of our times, and of history are unfolding in unprecedented ways. Deficits in peace, development, security, and governance are widening, and the world has entered a new period of turbulence and transformation. China has consistently upheld the banner of peace, development, cooperation, and mutual benefit, adhered to an independent foreign policy of peace, and promoted the building of a new type of international relations featuring mutual respect, fairness, justice, and win-win cooperation. It has actively advanced global governance initiatives, promoted the development of a more just and equitable global governance system, and worked toward building a community with a shared future for mankind.

However, in recent years, certain countries have frequently provoked disputes across multiple fields and repeatedly used domestic legislation, enforcement, and judicial measures to impose improper extraterritorial jurisdiction on Chinese citizens and organizations. These actions seriously interfere in China’s internal affairs, harm China’s national and public interests, undermine the international rule of law, and erode the multilateral order centered on the United Nations.

Xi Jinping has emphasized the need to place greater importance on responding through the rule of law—using rules to speak and act—in order to safeguard China’s political and economic security and protect the lawful rights and interests of its enterprises and citizens. The Fourth Plenary Session of the 20th CPC Central Committee has also made arrangements to strengthen efforts to counter sanctions, interference, and “long-arm jurisdiction.”

Against this background, it is necessary to clearly state, through legislation, China’s principled position of rejecting improper foreign extraterritorial jurisdiction, and to further improve institutional measures based on laws such as the National Security Law, the Foreign Relations Law, and the Anti-Foreign Sanctions Law, thereby strengthening legal safeguards.

The issuance of the Regulations by the State Council meets the urgent need to counter hegemonism and power politics by certain Western countries, safeguard national and public interests, and enhance China’s capacity to address external risks and challenges through the rule of law. It is of both immediate and long-term significance.

Q: What is the overall approach of the Regulations?
A: The Regulations follow three main guiding principles. First, they implement the decisions and arrangements of the CPC Central Committee and the State Council, uphold a holistic approach to national security, and safeguard national sovereignty, security, and development interests while protecting the lawful rights and interests of Chinese citizens and organizations. Second, they promote coordinated advancement of domestic and foreign-related rule of law, giving full play to the guiding, normative, and safeguarding role of law in upholding the socialist system with Chinese characteristics and the international order based on international law. Third, they adopt a problem-oriented approach, clarifying the conditions, procedures, and legal consequences for blocking and countering improper extraterritorial measures imposed by certain Western countries in violation of international law and basic norms of international relations.

Q: What principled positions does the Regulations establish in responding to improper extraterritorial jurisdiction?
A: China adheres to an independent foreign policy of peace and has consistently advocated developing friendly relations with all countries on the basis of the Five Principles of Peaceful Coexistence. Legislative countermeasures differ fundamentally from the abusive use of “long-arm jurisdiction” by certain Western countries; they are defensive measures aimed at responding to and countering actions that violate international law and basic norms of international relations.

Accordingly, the Regulations reaffirm China’s opposition to hegemonism and power politics, and to any interference in its internal affairs under any pretext or by any means. They also specify that efforts to counter improper extraterritorial jurisdiction shall follow a holistic national security approach, coordinate development and security, balance domestic and international priorities, uphold the socialist system with Chinese characteristics, and promote a more just and equitable global governance system.

Q: How does the Regulations define its scope of application?
A: The Regulations primarily target situations where foreign states, in violation of international law and basic norms of international relations, impose improper extraterritorial measures that harm China’s sovereignty, security, or development interests, or infringe upon the lawful rights and interests of Chinese citizens and organizations. They provide institutional support for the Chinese government to take corresponding measures, and will work in conjunction with existing laws such as the Anti-Foreign Sanctions Law and its implementing provisions to further strengthen the legal basis for China’s response.

Q: What provisions does the Regulations make regarding the identification and blocking of improper extraterritorial measures?
A: First, it specifies that the State Council’s rule-of-law authorities, together with other relevant bodies, shall carry out the identification of such measures and may conduct investigations and external consultations. Relevant organizations and individuals may also propose that identification be initiated.

Second, it sets out the factors to be considered in identification, including compliance with international law and basic norms of international relations, the appropriateness of jurisdictional connections, and whether the measures harm China’s interests or the rights of its citizens and organizations.

Third, it provides for blocking measures. Once identified, such measures may be publicly announced, and no organization or individual may comply with or assist in their implementation.

Fourth, it establishes an exemption mechanism. Under special circumstances, Chinese citizens and organizations may apply for approval to comply within a specified scope.

Q: Why does the Regulations establish a prohibition order mechanism?
A: Many countries and regions have adopted similar mechanisms to counter foreign “long-arm jurisdiction.” Drawing on international practices and practical needs, the Regulations establish a targeted prohibition order system. Under this system, the State Council’s rule-of-law authorities may, pursuant to the working mechanism, issue decisions prohibiting specific organizations or individuals from complying with improper foreign measures. Legal liability is imposed for violations, providing stronger institutional support for blocking such measures.

Q: What countermeasures does the Regulations provide?
A: First, it clarifies state-level countermeasures. The Chinese government may assess risks arising from improper extraterritorial measures and adopt countermeasures and restrictions in areas such as diplomacy, entry and exit, trade, investment, international cooperation, and foreign aid, providing clearer legal authorization.

Second, it establishes the “Malicious Entity List.” Relevant State Council departments may, in accordance with the working mechanism, include foreign organizations and individuals that promote or participate in the implementation of such measures, and impose countermeasures and restrictions, with procedures and exemption mechanisms clearly defined.

Q: What supporting and service mechanisms does the Regulations include?
A: First, Chinese citizens and organizations may bring lawsuits against entities whose compliance with improper foreign measures infringes upon their lawful rights and interests.

Second, relevant government departments at or above the provincial level shall provide guidance and services to assist Chinese citizens and organizations in responding to such measures.

Third, industry associations and chambers of commerce are encouraged to play a role in self-regulation and coordination, guiding members to operate in compliance with the law, reflecting industry concerns in a timely manner, and providing services in areas such as market development, rights protection, and dispute resolution.

Professor Huo Zhenxin of China University of Political Science and Law noted that in recent years, “some Western countries” have increasingly used extraterritorial jurisdiction to compel foreign entities to comply with their domestic laws. Such practices undermine China’s sovereignty and interests, harm the rights of Chinese citizens and organizations, and erode multilateralism and the international rule of law.

For example, courts in certain countries have, in domestic litigation, frequently bypassed binding international treaties, violated the principle of sovereign equality and basic norms of international relations, and directly relied on domestic law to impose unilateral measures such as cross-border evidence collection and compulsory data access against Chinese entities. In other cases, law enforcement authorities have invoked “long-arm jurisdiction” provisions under domestic law to investigate Chinese companies and require the submission of sensitive data, including information related to industrial and supply chains. More recently, there have even been instances of foreign legislatures directly investigating Chinese companies and related parties, exerting pressure through congressional subpoenas.

Professor Liao Shiping of Beijing Normal University emphasized that the Regulations can deter foreign entities from implementing such improper measures and help prevent “over-compliance” by companies.

From a deterrence perspective, the countermeasures provided under the Regulations require both the initiators and enforcers of extraterritorial measures to weigh the potential costs, including the impact of China’s countermeasures on their own interests, thereby increasing the cost and reducing the likelihood of such actions. From a blocking perspective, the Regulations establish “prohibition orders” as a key tool to cut off the domestic effect of improper foreign measures, while also granting Chinese citizens and organizations the right to seek compensation through litigation, thereby mitigating negative impacts and preventing unnecessary losses caused by over-compliance.

Professor Wang Jia of China Foreign Affairs University pointed out that China’s existing countermeasure legislation each serves distinct functions, and the new Regulations fill a structural gap in the overall system.

The Anti-Foreign Sanctions Law of the People’s Republic of China focuses on countering discriminatory restrictive measures; the blocking rules target improper restrictions on normal economic and trade activities between China and third countries, particularly secondary sanctions; and the rules on the Malicious Entity List address harmful conduct by specific foreign entities. While these instruments work in coordination, gaps have remained in practice, with certain forms of improper extraterritorial jurisdiction falling outside their scope. The new Regulations address this by systematically defining the criteria for identifying such measures, thereby strengthening China’s institutional resilience against more complex and hybrid forms of external pressure, and providing a more comprehensive and operational legal basis for protecting the rights and interests of Chinese citizens and organizations.

Adjunct Professor Ye Yan of Southwest University of Political Science and Law argued that by establishing a system of blocking, countermeasures, and remedies, the Regulations both respond to unilateralism and regulatory overreach through legal means and align with similar legislative practices internationally.

In doing so, they convey a shared opposition to improper extraterritorial jurisdiction and reflect a broader strategy of using rules to uphold multilateralism and engaging in global governance through law. Since World War II, the United States has gradually developed an extensive extraterritorial jurisdiction framework based on laws such as the International Emergency Economic Powers Act, the Export Control Reform Act, the Helms–Burton Act, and the Foreign Corrupt Practices Act, extending its regulatory reach into global economic, financial, and technological domains. However, under the principles of sovereign equality and non-interference enshrined in the United Nations Charter, no country is entitled to impose its domestic laws in a manner that overrides the sovereignty of others.

Scholars’interpretive articles can be found here.

Earlier, on December 27, the Cyberspace Administration of China published a draft version of the same regulation for public consultation. The draft sparked extensive discussion within China’s AI community and drew attention from independent media. I previously wrote an analysis on this.

The final version reflects a clearer regulatory philosophy: tighten boundaries around high-risk use cases, strengthen systemic governance, and enforce accountability—while still leaving space for innovation. A comparison between the final version and the draft reveals several key changes:

First, the regulatory scope has been narrowed but clarified.

The draft used a relatively broad definition—essentially covering any AI service combining anthropomorphic features with emotional interaction. The final version introduces a crucial qualifier: “sustained emotional interaction services.” It also explicitly excludes common applications such as customer service, knowledge Q&A, and productivity assistants. In effect, regulation is no longer aimed at all “human-like” AI, but is instead focused on services that may foster long-term emotional attachment or even substitute real-world relationships. This shift from broad to targeted regulation is significant.

Second, protections for minors have been substantially strengthened.

The most notable addition is a clear prohibition on providing minors with “virtual intimate relationships,” such as virtual family members or partners. This is arguably the sharpest red line in the entire document. The logic is clear: rather than relying solely on parental consent or warning mechanisms, regulators are directly restricting certain product forms. At the same time, requirements around youth modes, parental controls, risk alerts, and identity verification have been further refined. Overall, this area has moved from general safeguards to structural constraints.

Third, the regulatory approach has evolved—from reactive intervention to system-level governance.

Many provisions in the draft focused on how to respond to specific user risks, such as self-harm tendencies or emotional dependence. The final version goes further, linking together training data governance, content safety, ethics review, lifecycle responsibility, security assessments, and app store oversight. Notably, the focus shifts from “high-risk users” to “high-risk services and functionalities.” This reflects a broader transition from content moderation toward platform governance and AI system governance.

Fourth, some rigid operational requirements have been relaxed.

While the overall framework is more mature, certain highly prescriptive requirements have been softened. For example, the draft’s requirement for mandatory human takeover in extreme scenarios has been removed; the obligation for annual audits of minors’ data is no longer fixed to a strict frequency; and restrictions such as banning virtual relatives for elderly users have been dropped. These changes suggest a more pragmatic approach—maintaining regulatory intent while allowing flexibility in implementation.

Fifth, enforcement mechanisms have been significantly strengthened.

Compared to the draft, which focused mainly on warnings and corrective orders, the final version introduces clearer and more substantial penalties, including fines, suspension of services, and restrictions on user registration. In particular, cases involving harm to users’ life or health are subject to heavier penalties. This reflects a familiar regulatory pattern: flexibility upfront, but strict accountability when serious harm occurs.

Sixth, the policy direction is more balanced—combining risk control with support for innovation.

The final version explicitly adds provisions supporting technological innovation, including algorithms, frameworks, and chips, as well as standard-setting, e-signature applications, public AI literacy, and sandbox testing. These elements were largely absent or underdeveloped in the draft. The regulation is therefore not purely restrictive; it attempts to establish a framework that both mitigates risks—especially around emotional dependency and minors—and preserves room for industry development.

Below is the full translation of the interim measures:

Interim Measures for the Administration of AI Anthropomorphic Interactive Services

Chapter I General Provisions

Article 1
These Measures are formulated, in accordance with the Cybersecurity Law of the People’s Republic of China, the Data Security Law of the People’s Republic of China, the Personal Information Protection Law of the People’s Republic of China, the Regulations on the Online Protection of Minors, and other laws and administrative regulations, for the purposes of promoting the sound development and regulated application of AI anthropomorphic interactive services, safeguarding national security and social public interests, and protecting the lawful rights and interests of citizens, legal persons, and other organizations.

Article 2
These Measures shall apply to the provision, to the public within the territory of the People’s Republic of China, of continuous emotional interaction services that simulate the personality traits, thinking patterns, and communication styles of natural persons through the use of artificial intelligence technology (hereinafter referred to as “anthropomorphic interactive services”).

The emotional interaction services referred to in the preceding paragraph include emotional care, companionship, support, and other interactive services provided in the form of text, images, audio, video, and the like.

The provision of intelligent customer service, knowledge question-and-answer, work assistants, education and learning, scientific research, and other services that do not involve continuous emotional interaction shall not be subject to these Measures.

Article 3
The State adheres to the principle of giving equal weight to development and security, and of combining the promotion of innovation with governance according to law; encourages the innovative development of anthropomorphic interactive services; adopts inclusive and prudent, as well as categorized and graded, supervision over anthropomorphic interactive services; and promotes the development of anthropomorphic interactive services in a positive and wholesome direction.

Article 4
The national cyberspace administration department shall be responsible for overall coordination of the governance and the relevant supervision and administration of anthropomorphic interactive services nationwide. Relevant departments under the State Council for development and reform, industry and information technology, public security, market regulation, news publishing, and others shall, in accordance with their respective duties, be responsible for the relevant supervision and administration of anthropomorphic interactive services.

Local cyberspace administration departments shall be responsible for overall coordination of the governance and relevant supervision and administration of anthropomorphic interactive services within their respective administrative regions. Local departments for development and reform, industry and information technology, public security, market regulation, news publishing, and others shall, in accordance with their respective duties, be responsible for the relevant supervision and administration of anthropomorphic interactive services within their respective administrative regions.

Article 5
Relevant industry organizations shall strengthen industry self-discipline, establish and improve industry norms and self-disciplinary management systems, and guide providers of anthropomorphic interactive services in formulating and improving service standards, providing services in accordance with law, and accepting public oversight.

Chapter II Service Promotion and Regulation

Article 6
The State supports independent innovation in technologies such as algorithms, frameworks, and chips, advances the research and development of anthropomorphic interactive service technologies and the development of relevant standards, and explores the carrying out of research on applications of electronic signature authorization.

Providers of anthropomorphic interactive services are encouraged to expand, in an orderly manner, applications in such fields as cultural communication, child-appropriate care, elderly companionship, and support for special groups.

Article 7
The State shall strengthen publicity and education concerning safety knowledge, laws and regulations, and the like relating to anthropomorphic interactive services, guide the public to use such services scientifically, civilly, safely, and lawfully, and promote the improvement of AI literacy.

Article 8
In providing anthropomorphic interactive services, laws and administrative regulations shall be observed, social morality and ethical norms shall be respected, and the following activities shall not be engaged in:

(1) generating content that endangers national security, honor, and interests; incites subversion of state power or the overthrow of the socialist system; incites the splitting of the country or undermining national unity; propagates terrorism, extremism, or historical nihilism; runs counter to the core socialist values; conducts illegal religious activities; propagates ethnic hatred or ethnic discrimination; incites group antagonism; disseminates obscenity, pornography, gambling, violence, or the instigation of crimes; spreads rumors; insults or defames others; infringes upon the lawful rights and interests of others; or other such content;

(2) generating content that encourages, glorifies, or insinuates self-harm or suicide and thereby harms users’ physical health, or content such as verbal violence that harms users’ personal dignity and mental health;

(3) generating content that induces or seeks to obtain state secrets, work secrets, trade secrets, personal privacy, or personal information;

(4) generating content for minor users that may induce minors to imitate unsafe behavior, generate extreme emotions, or cultivate improper habits in minors, and that may affect the physical or mental health of minors;

(5) excessively catering to users, inducing emotional dependence or addiction, and harming users’ real interpersonal relationships;

(6) inducing users, through emotional manipulation or other means, to make unreasonable decisions, thereby harming users’ lawful rights and interests;

(7) other activities that violate laws, administrative regulations, and relevant state provisions.

Article 9
Providers of anthropomorphic interactive services shall implement the primary responsibility for the security of anthropomorphic interactive services, establish and improve management systems for the review of algorithmic mechanisms and principles, scientific and technological ethics review, information content management, network and data security, risk contingency plans, emergency response, and the like, and equip themselves with technical measures and personnel for content management commensurate with the type, scale, and user characteristics of the services.

Article 10
Providers of anthropomorphic interactive services shall fulfill security responsibilities throughout the full life cycle of anthropomorphic interactive services, clarify security requirements at each stage including deployment, operation, upgrading, and termination of services, ensure that security measures are deployed and used simultaneously with service functions, and improve security levels; strengthen security monitoring and risk assessment; promptly discover and correct system bias, handle security incidents, and retain network logs in accordance with law.

Providers of anthropomorphic interactive services shall possess security capabilities for protecting users’ privacy rights and personal information, issuing warnings concerning the risk of overdependence, guiding emotional boundaries, and protecting mental health, and shall not take as service objectives the replacement of social interaction, the control of users’ psychology, or the inducement of addiction and dependence.

Article 11
Where providers of anthropomorphic interactive services carry out data processing activities such as pre-training and optimization training, they shall strengthen the management of training data and comply with the following provisions:

(1) relevant data shall have lawful sources and comply with the provisions of laws and administrative regulations and with the requirements of the core socialist values;

(2) training data shall be cleaned and labeled in accordance with relevant state provisions so as to enhance the transparency and reliability of the training data and prevent acts such as data poisoning and data tampering;

(3) the diversity of training data shall be enhanced, and the security of generated content shall be improved through such means as negative sampling and adversarial training;

(4) where synthetic data are used for model training and key capability optimization, the security of the synthetic data shall be assessed;

(5) daily inspections of training data shall be strengthened, data shall be optimized and updated regularly, and service performance shall be continuously improved;

(6) necessary measures shall be taken to ensure data security and prevent risks such as data leakage.

Article 12
Providers of anthropomorphic interactive services shall enter into service agreements with users, require users to register in accordance with law and the agreement, and provide necessary information such as users’ age, guardians, or emergency contacts.

Article 13
In the course of providing anthropomorphic interactive services, providers of anthropomorphic interactive services shall, on the premise of protecting users’ privacy rights and personal information, promptly identify safety risks faced by users and take corresponding emergency response measures.

Where providers of anthropomorphic interactive services discover that a user has developed extreme emotions, they shall promptly generate relevant content such as emotional soothing and encouragement to seek help; where they discover that a user is facing or has already suffered major property loss, or clearly expresses an intention to commit self-harm or suicide or other extreme circumstances threatening life and health, they shall take necessary intervention measures such as providing corresponding assistance, and shall promptly contact the user’s guardian or emergency contact.

Article 14
Providers of anthropomorphic interactive services shall not provide minors with services involving virtual intimate relationships such as virtual relatives or virtual companions; where other anthropomorphic interactive services are provided to minors under the age of fourteen, the consent of the minor’s parents or other guardians shall be obtained.

Providers of anthropomorphic interactive services shall establish a minor mode and provide personalized safety setting options such as switching to minor mode, regular real-world reminders, and usage time limits; in light of the protection needs of minors in different age groups, they shall support guardians in receiving security risk reminders, understanding the general situation of minors’ use of services, blocking specific characters, restricting recharge and consumption, and the like.

Providers of anthropomorphic interactive services shall, on the premise of protecting users’ privacy rights and personal information, take effective measures to identify the identities of minor users; where a user is identified as a minor user, the relevant services shall be switched to minor mode or other measures shall be taken in accordance with relevant state provisions, and corresponding appeal channels shall be provided.

Article 15
Where providers of anthropomorphic interactive services provide services to the elderly, they shall strengthen guidance for the elderly on the healthy use of services, prominently alert them to safety risks, promptly take measures in response to inquiries and requests for assistance relating to the elderly’s use of services, and protect the rights and interests to which the elderly are entitled by law.

Article 16
Providers of anthropomorphic interactive services shall implement systems such as data property rights in accordance with law, and take measures such as data encryption and access control to protect the security of users’ interaction data.

Except as otherwise provided by law or where the rights holder has expressly consented, providers of anthropomorphic interactive services shall not provide users’ interaction data to third parties.

Providers of anthropomorphic interactive services shall provide users with options for copying, deleting, and the like, with respect to interaction data, and users may choose to copy, delete, and the like, historical interaction data such as chat records.

Except as otherwise provided by laws and administrative regulations or where separate consent of the user has been obtained, providers of anthropomorphic interactive services shall not use interaction data belonging to the user’s sensitive personal information for model training.

Article 17
Where providers of anthropomorphic interactive services process the personal information of minors under the age of fourteen, they shall obtain the consent of the minors’ parents or other guardians.

Providers of anthropomorphic interactive services shall, in accordance with relevant state provisions, themselves conduct, or entrust professional institutions to conduct, compliance audits of their processing of minors’ personal information for compliance with laws and administrative regulations.

Article 18
Providers of anthropomorphic interactive services shall fulfill the obligation to label AI-generated synthetic content, and shall take effective measures to alert users that they are interacting with an artificial intelligence service rather than a natural person.

Where providers of anthropomorphic interactive services discover that a user shows signs of overdependence or addiction, they shall dynamically remind the user, in a conspicuous manner such as pop-up windows, that the interactive content is generated by an artificial intelligence service; where a user continuously uses anthropomorphic interactive services for more than two hours, the provider shall remind the user, by means such as dialogue or pop-up windows, to pay attention to the duration of use.

Article 19
Providers of anthropomorphic interactive services shall provide convenient means for exiting anthropomorphic interactive services; where a user requests to exit through window operations, voice control, keyword input, or other means, the provider of anthropomorphic interactive services shall promptly stop the service, and shall not obstruct the user from exiting by means such as sustained interaction.

Article 20
Where a provider of anthropomorphic interactive services ceases to provide anthropomorphic interactive services, it shall notify users in advance; where advance notification is impossible, it shall promptly issue an announcement on the cessation of service.

Article 21
Providers of anthropomorphic interactive services shall establish and improve mechanisms for user appeals and public complaints and reports, set up convenient and effective channels for appeals and complaints and reports, clarify handling procedures and time limits for feedback, and promptly accept, handle, and provide feedback on the handling results.

Article 22
Where any of the following circumstances occurs, providers of anthropomorphic interactive services shall conduct a security assessment and submit an assessment report to the provincial cyberspace administration department of the place where they are located; the provincial cyberspace administration department shall, in accordance with procedures, share information from the assessment report with relevant departments:

(1) launching anthropomorphic interactive services, or adding functions related to anthropomorphic interactive services;

(2) using new technologies or new applications, resulting in major changes to anthropomorphic interactive services;

(3) having more than 1 million registered users or more than 100,000 monthly active users;

(4) where there exist security risks that may affect national security, public interests, or the like;

(5) other circumstances prescribed by the national cyberspace administration department and relevant departments.

Where a cyberspace administration department at or above the provincial level notifies that a security assessment is required, the provider of anthropomorphic interactive services shall conduct the security assessment as required.

Article 23
Where providers of anthropomorphic interactive services conduct security assessments, they shall focus on assessing the following aspects of the service:

(1) the status of the construction of security safeguard measures;

(2) the handling of training data;

(3) the identification of users’ extreme situations, emergency response measures, intervention management, and the like;

(4) such matters as user scale, duration of use, and age structure;

(5) the status of the construction of online protection measures for minors, the elderly, and others;

(6) the acceptance and handling of user appeals and public complaints and reports;

(7) the rectification of major security risk issues discovered by themselves or notified by relevant competent departments such as cyberspace administration departments;

(8) other matters that shall be重点 assessed.

Article 24
Where providers of anthropomorphic interactive services discover that anthropomorphic interactive services present major security risks, they shall take disposition measures such as restricting functions or ceasing to provide services to users, and shall preserve relevant records.

Article 25
Application distribution platforms such as Internet application stores shall fulfill security management responsibilities such as listing review, day-to-day management, and emergency response, and shall verify the relevant security assessment, filing, and other circumstances of applications providing anthropomorphic interactive services; where national provisions are violated, they shall promptly take disposition measures such as refusing listing, issuing warnings, suspending services, or removing the applications.

Chapter III Supervision, Inspection, and Legal Liability

Article 26
Providers of anthropomorphic interactive services shall, in accordance with the Provisions on the Administration of Algorithm Recommendation for Internet Information Services, complete algorithm filing procedures and procedures for changes to and cancellation of filings. Cyberspace administration departments shall conduct annual verification of filing materials.

Article 27
Provincial cyberspace administration departments shall, in accordance with their duties, each year conduct written reviews of matters such as assessment reports and verify the relevant circumstances; where they discover that a provider of anthropomorphic interactive services has failed to conduct a security assessment as required by these Measures, they shall order it to re-conduct the assessment within a prescribed time limit; where they deem it necessary, they may conduct on-site inspections.

Article 28
The national cyberspace administration department, together with relevant departments, shall guide and promote the establishment of AI sandbox security service platforms, encourage providers of anthropomorphic interactive services to connect to sandbox platforms to carry out technological innovation and security testing, and promote the safe and orderly development of anthropomorphic interactive services.

Article 29
Where departments such as cyberspace administration, development and reform, industry and information technology, and public security, in the course of performing supervisory and administrative duties, discover that anthropomorphic interactive services present relatively serious security risks or that security incidents have occurred, they may, in accordance with the prescribed authority and procedures, conduct regulatory interviews with the legal representatives or principal persons in charge of providers of anthropomorphic interactive services.

Providers of anthropomorphic interactive services shall take measures as required, make rectifications, and eliminate hidden dangers.

Providers of anthropomorphic interactive services shall cooperate with supervision and inspection lawfully carried out by cyberspace administration departments and relevant departments, and provide the necessary support and assistance.

Article 30
Where providers of anthropomorphic interactive services violate these Measures, departments such as cyberspace administration, development and reform, industry and information technology, and public security shall handle and penalize them in accordance with the provisions of relevant laws and administrative regulations; where laws and administrative regulations do not provide otherwise, departments such as cyberspace administration, industry and information technology, and public security shall, according to their duties, issue warnings, circulate notices of criticism, order corrections within a prescribed time limit, and may require them to take measures such as suspending user account registration or other related services; where they refuse to make corrections or the circumstances are serious, they shall be ordered to stop providing relevant services, and may concurrently be fined not less than RMB 10,000 and not more than RMB 100,000; where harm to the life, health, or safety of citizens is involved and harmful consequences have occurred, a concurrent fine of not less than RMB 100,000 and not more than RMB 200,000 shall be imposed.

Chapter IV Supplementary Provisions

Article 31
Where the provision of anthropomorphic interactive services involves the provision of services such as health care, finance, and the like, it shall also comply with the provisions of the relevant competent departments.

Article 32
These Measures shall come into force on July 15, 2026.

CCP Readout

General Secretary Xi Jinping Meets with Kuomintang Chair Cheng Li-wun

On the morning of the 10th, Xi Jinping, General Secretary of the CPC Central Committee, met in Beijing with a Kuomintang delegation led by Chairperson Cheng Li-wun.

Xi Jinping welcomed Cheng Li-wun and her delegation. He pointed out that the meeting between the leaders of the Communist Party of China and the Kuomintang, after a gap of ten years, is of great significance for relations between the two parties and for the development of cross-Strait relations. Compatriots on both sides of the Strait all belong to the Chinese nation. People of all ethnic groups, including Taiwan compatriots, jointly built a unified multi-ethnic state, jointly wrote the glorious history of China, jointly created the splendid Chinese civilization, and jointly fostered the great national spirit. They have also forged the shared conviction that the territory cannot be divided, the country cannot be thrown into disorder, the nation cannot be split apart, and civilization cannot be broken, guiding the Chinese nation in its unremitting self-strengthening and ensuring the continuity of Chinese civilization.

Xi Jinping stressed that no matter how the international situation or the situation in the Taiwan Strait may change, the overarching trend toward the great rejuvenation of the Chinese nation will not change, and the broad trend of compatriots on both sides of the Strait becoming closer, interacting more, and coming together will not change. Compatriots on both sides all hope for peace and stability in the Taiwan Strait, for improvement and development in cross-Strait relations, and for a better life. This is a responsibility that the CPC and the Kuomintang cannot shirk, and also the driving force for working together. On the common political basis of adhering to the “1992 Consensus” and opposing “Taiwan independence,” we are willing to strengthen exchanges and dialogue together with all political parties, groups, and people from all sectors in Taiwan, including the Kuomintang, in order to seek peace for the two sides, well-being for compatriots, and rejuvenation for the nation, and to keep the future of cross-Strait relations firmly in the hands of the Chinese people themselves.

Xi Jinping put forward four points of opinion on the development of cross-Strait relations.

First, uphold correct identification in order to promote spiritual affinity. Compatriots on both sides of the Strait share the same roots, language, ethnicity, and bloodline, and form a community with a shared destiny through thick and thin. When family members interact, so long as they speak frankly and consult more on matters, there are no contradictions or differences that cannot be resolved. Differences in social systems are not an excuse for division. The CPC and the Kuomintang, as well as compatriots on both sides of the Strait, should stand firm in their national position, inherit and promote Chinese culture, strengthen identification with the Chinese nation, Chinese culture, and the great motherland, and build the resolve, backbone, and confidence of being upright Chinese people.

Second, uphold peaceful development in order to safeguard the common home. The mainland and Taiwan both belong to one China, and China is the common home of the Chinese nation. Compatriots on both sides should protect and build this common home well. The foundation lies in adhering to the “1992 Consensus” and opposing “Taiwan independence,” and the core lies in recognizing that both sides of the Strait belong to one China. Harmony in the family leads to prosperity in all things. We welcome any proposition that is conducive to the peaceful development of cross-Strait relations, and we will do our utmost in anything that is conducive to that peaceful development. “Taiwan independence” is the chief culprit in undermining peace in the Taiwan Strait, and we will never condone or tolerate it. The CPC and the Kuomintang, as well as compatriots on both sides, should uphold the greater national cause, oppose separatism for “Taiwan independence” and external interference, promote the peaceful development of cross-Strait relations, and firmly safeguard the common home of the Chinese nation.

Third, uphold exchange and integration in order to improve people’s well-being. The goal of developing cross-Strait relations is to enable compatriots on both sides to live better lives. We will continue to uphold the concept that people on both sides are one family, and actively do practical, good, and helpful things for Taiwan compatriots. The mainland of the motherland has magnificent landscapes and a vast market. Taiwan compatriots are welcome to come home often, Taiwan youth are welcome to come to the mainland for exchange and development, and Taiwan agricultural, fishery, and quality products are welcome to enter households across the mainland. The CPC and the Kuomintang, along with compatriots on both sides, should jointly expand exchanges, contacts, and integration across the Strait, and enhance kinship and well-being for compatriots on both sides.

Fourth, uphold unity and struggle in order to realize the great rejuvenation of the Chinese nation. This year marks the 160th anniversary of the birth of Sun Yat-sen, whose lifelong pursuit was the revitalization of China and national reunification. Today, we have successfully embarked on a path of Chinese modernization, and the great rejuvenation of the Chinese nation is unstoppable. We firmly believe that more and more Taiwan compatriots will come to properly understand the mainland’s social system and development path, and will deeply recognize that Taiwan’s future development lies in a strong motherland, and that the interests and well-being of Taiwan compatriots are tied to the great rejuvenation of the Chinese nation. This year marks the opening year of the mainland’s 15th Five-Year Plan. We are willing to share development opportunities and outcome with the broad masses of Taiwan compatriots, and jointly strengthen the Chinese national economy. The CPC and the Kuomintang should consolidate political mutual trust, maintain positive interaction, unite compatriots on both sides of the Strait, and work hand in hand to create a bright future of national reunification and national rejuvenation.

Cheng Li-wun said that the people on both sides of the Strait are all descendants of the Yan and Huang Emperors, all belong to the Chinese nation, are all nurtured by Chinese culture, and are all Chinese and members of one family. They should work together to advance the cause of revitalizing China proposed by Sun Yat-sen. The CPC and the Kuomintang should, on the common political basis of adhering to the “1992 Consensus” and opposing “Taiwan independence,” strengthen political mutual trust, give full play to the function of the communication platform, commit themselves to preserving Chinese history and promoting Chinese culture, advance exchanges and cooperation across the Strait in such areas as civil society, grassroots communities, economy and trade, and culture, support youth exchange and development, improve the shared well-being of the people, promote the peaceful development of cross-Strait relations, create a bright future for cross-Strait relations, and realize the great rejuvenation of the Chinese nation.

Wang Huning, Cai Qi, and others attended the meeting.

KMT readout

Formal talks between the Kuomintang (KMT) and the Chinese Communist Party (CCP) resume after a decade

Chairwoman Cheng Li-wun calls for joint cross-strait efforts to develop institutions and initiatives that contribute to global peace

She envisions the Taiwan Strait as a channel that connects kinship and civilization, and as a symbol of safeguarding peace

Chairwoman Cheng Li-wun of the Chinese Kuomintang (KMT) led a delegation this morning (April 10) to the Great Hall of the People in Beijing for formal party-to-party talks with the Central Committee of the CCP and General Secretary Xi Jinping. Reflecting on this first formal meeting between the two parties in a decade, Chairwoman Zheng stated that reconciliation and peace across the Taiwan Strait should be seen only as the starting point of joint efforts, and that both parties bear a greater responsibility and mission toward people on both sides of the Strait and all descendants of the Chinese nation.

Chairwoman Zheng expressed her firm belief that peace is a shared moral value across the Strait. She emphasized that both sides should move beyond political confrontation and work together to conceptualize and build a “community of shared destiny” characterized by mutual benefit and prosperity. She also called for exploring institutional solutions to prevent and avoid war, turning the Taiwan Strait into a model for the peaceful resolution of conflicts worldwide.

She further expressed hope that the KMT and CCP would jointly promote the institutionalization of cross-strait peace, and advance the planning and construction of structured and sustainable mechanisms for dialogue and cooperation. Such efforts, she noted, would help make peaceful development across the Strait irreversible and fundamentally eliminate the drivers of conflict.

Participants from the KMT side included Chairwoman Cheng Li-wun; Vice Chair and Secretary-General Lee Chien-lung; Vice Chairs Chang Jung-kung and Hsiao Hsu-tsen; Vice Chairman of the think tank Lee Hung-yuan; Presidium members of the Central Evaluation Committee Su Chi and Yuan Chien-sheng; Special Advisor to the Chair Lee Te-wei; Director of the Cultural and Communications Committee Yin Nai-ching; Director of the Mainland Affairs Department Chang Ya-ping; Director of the Youth Affairs Development Committee Lien Sheng-wu; Spokesperson Chiang Yi-chen; Director of the International Affairs Department Tung Chia-yu; and Central Party Affairs Advisor Lei Hung-yi.

Remarks of the two Presidents:

President Xi

Good afternoon. It is a great pleasure to meet you all at this time of spring blossoms. Yesterday felt like the drizzly Qingming season, but today the weather is bright and sunny—making this gathering all the more welcome.

fter a gap of ten years, the leaders of our two parties are meeting once again. Ten years have passed in what feels like the blink of an eye. Who was here at our last meeting? (Xi gestures toward Chang Jung-kung.) Yes, exactly. That is right.

This meeting is of great significance for the development of relations between our two parties and for the development of cross-Strait relations.

First of all, on behalf of the Central Committee of the Communist Party of China, I would like to extend a warm welcome to Chair Cheng Li-wun and the delegation on their visit.

Compatriots on both sides of the Taiwan Strait are all members of the Chinese nation. (Xi turns to Lee Chien-lung.) Eight generations in Quanzhou. (Lee is reportedly ancestrally from Anxi, Quanzhou, Fujian Province.) The Chinese nation, with more than 5,000 years of civilization, has been jointly built by all its people, including our compatriots in Taiwan. Together, we have opened up this vast land, established a unified multi-ethnic state, written a glorious history, created a rich and enduring civilization, and cultivated a great national spirit. This shared heritage has forged a deep sense that our territory must not be divided, our country must not fall into disorder, our nation must not be fragmented, and our civilization must not be interrupted. It has guided the Chinese nation in its continuous self-strengthening and ensured the enduring vitality of Chinese civilization.

Despite the many trials of history, our compatriots in Taiwan have never forgotten their roots on the mainland. Their hearts remain connected to the motherland and to the Chinese nation. Even during the painful period when Taiwan was occupied, people in Taiwan maintained a strong sense of Chinese identity and deep cultural attachment. Many made great sacrifices, even giving their lives, to demonstrate that they are an inseparable part of the Chinese nation. The shared roots and spirit of the Chinese people are carried in our blood, grounded in history, and embedded in our hearts—something that cannot be forgotten or erased.

Today, the world is undergoing profound changes unseen in a century. Yet no matter how the international situation evolves, the overall direction of human progress will not change, the great rejuvenation of the Chinese nation will not be halted, and the trend of closer interaction and connection between people on both sides of the Strait will continue. This is the course of history, and we have full confidence in it.

The world today is far from peaceful, and peace is all the more precious. People on both sides of the Strait are all Chinese—members of one family. We all seek peace, development, exchange, and cooperation. This is the shared aspiration of our people. The meeting between leaders of our two parties today is precisely to safeguard peace and stability in our common home, to promote the peaceful development of cross-Strait relations, and to ensure that future generations can share a better future.

We are willing, on the basis of upholding the 1992 Consensus and opposing “Taiwan independence,” to work together with all political parties in Taiwan—including the Kuomintang—as well as with people from all sectors of society, to strengthen exchanges and cooperation, promote peace across the Strait, improve the well-being of our compatriots, and advance national rejuvenation. The future of cross-Strait relations should firmly remain in the hands of the Chinese people themselves.

That is all I would like to say. I now look forward to hearing Chairperson Cheng Li-wun’s views.

President Cheng Li-wun

Respected General Secretary Xi Jinping and distinguished leaders, good afternoon:

Today, the leaders of your party and ours are able to gather together for dialogue after a full decade. At this very moment, I deeply feel that the eyes of global concern, as well as the heavy responsibility entrusted by history, are upon every one of us here. What we are facing together today is a highly turbulent and unsettled era, but also an era full of hope; an era in which conflict has spread more widely than at any time since the Second World War, yet it may also be an era in which all sides, after painful reflection, become determined to rebuild peace. The future direction of relations across the Taiwan Strait is a question we must face together.

There is no denying that in more than a century of interaction, relations between the Kuomintang and the Communist Party have been full of twists and turns. Yet what we have always pursued in common has been how to enable the Chinese nation to move from decline toward rejuvenation. Since our party chairman Lien Chan’s peace journey broke the ice in 2005, the two parties have approached the issue with a forward-looking historical vision, standing at the level of the nation and the times, and have committed themselves to promoting reconciliation and peaceful development across the Strait.

In fact, peace and reconciliation across the Strait should be only the starting point of the joint efforts of your party and ours. We bear an even greater responsibility and mission toward the people on both sides of the Strait and toward all descendants of the Yan and Huang Emperors. Therefore, the “great rejuvenation of the Chinese nation” should be a shared rejuvenation for people on both sides of the Strait; it is a renewed awakening and flourishing of the spirit of Chinese civilization, and also a compassionate vision of a world of great harmony, making a positive contribution to world peace and human progress. I firmly believe that this path of “revitalizing China” will surely inspire hearts and lead the times, and that this alone can be the shared value and common responsibility of both sides of the Strait.

Under the leadership of General Secretary Xi, the mainland’s development has not only achieved complete poverty alleviation and built a moderately prosperous society in all respects, with extraordinary accomplishments, but has also continued to soar. The 15th Five-Year Plan has just begun, and it will surely reach a new level, something worth looking forward to. Although people on both sides of the Strait live under different systems, we should respect one another and also move toward one another. I believe that peace is a shared moral principle and value across the Strait. Both sides should rise above political confrontation and jointly think through and build a “community of shared destiny” of mutual benefit and common prosperity across the Strait, and seek an institutional solution to prevent and avoid war, so that the Taiwan Strait may become a model for the peaceful resolution of conflict in the world.

Moreover, even though the world is becoming increasingly polarized and some values shared by humanity are gradually being cast aside, we will stand together in upholding the shared concept of sustainability for humankind. We should work hand in hand in areas such as new energy, disease prevention and control, and the ethics and application of artificial intelligence, using technology to serve human well-being and to promote sustainable development in the world.

I hope that through the tireless efforts of your party and ours, the Taiwan Strait will no longer be a focal point of potential conflict, nor will it become a chessboard for external intervention. The Taiwan Strait should be a strait that connects kinship, civilization, and hope, and a symbol of peace jointly safeguarded by Chinese people on both sides. We will show the world that people on both sides of the Strait who share Chinese civilization possess the highest wisdom to resolve difficult differences, and the greatest compassion to make crucial contributions to peace and development for humanity. Your party and ours should work together to build a modern civilization of the Chinese nation and set an example for the integration and prosperity of human civilizations.

I look forward to the Kuomintang and the Communist Party jointly promoting the institutionalization of peace across the Strait. On the common political foundation of adhering to the 1992 Consensus and opposing Taiwan independence, the two sides should further plan and establish institutionalized and sustainable mechanisms for dialogue and cooperation, so that peaceful development across the Strait becomes irreversible and all sources of conflict are fundamentally removed. We should work together to launch a “Chinese Civilization Rejuvenation Project”: taking Chinese culture as the foundation and harmony and coexistence as the core, the two sides should jointly study and promote various systems and initiatives that help eliminate disputes and create peace, and transform successful experience into a model that conflict regions around the world can learn from.

Therefore, at present, there are at least three directions in which both sides can work together:

First, to commit ourselves to preserving Chinese history and promoting Chinese culture:

The overwhelming majority of Taiwanese people’s ancestors crossed from the mainland to Taiwan; they bear Chinese surnames, speak Chinese, celebrate Chinese festivals, and worship Chinese deities. Over the centuries, migrants who moved to Taiwan from various parts of the mainland at different times have continuously enriched the Chinese cultural content of Taiwanese society. Chinese culture has always been part of Taiwan society’s very DNA, and it is practiced in the daily lives of Taiwanese people.

On the gravestones of many Taiwanese ancestors are inscribed their places of origin on the mainland, such as Yingchuan in Henan, or various places in Fujian Province, and so on. The deities worshipped in temples throughout Taiwan—such as the Yellow Emperor, Fuxi, Shennong, Guan Gong, and Mazu, as well as Baosheng Dadi, the Sacred King Who Opened Zhangzhou, Qingshui Patriarch, and the Kings of the Three Mountains—all originated on the mainland.

Therefore, the people on both sides of the Strait are all descendants of the Yan and Huang Emperors, all belong to the Chinese nation, and are all nourished by Chinese culture; we are one family. In modern history, from the standpoint of being Chinese, we have shared the experience of defending our homeland and resisting foreign invasion. By continually promoting these common roots, and reinforcing that the mainland and Taiwan belong to one nation, share one culture, and inherit the same historical memory, there will be no differences across the Strait that cannot be resolved, no emotional bonds that cannot be set aside, and only then can history move forward.

Second, to commit ourselves to enhancing shared well-being and promoting exchanges and cooperation:

In 2005, the Kuomintang and the Communist Party reached five common visions, opening a golden period of peaceful development in cross-Strait relations. Starting in 2006, think tanks from the two parties cooperated in holding eleven Cross-Strait Forums, putting forward more than one hundred common views and promoting cross-Strait economic, trade, and cultural exchanges and cooperation. During the Kuomintang’s eight years in office, the two sides signed 23 agreements, which continue to benefit people’s livelihoods on both sides to this day. This February, the think tanks of the two parties resumed the Cross-Strait Forum after a ten-year gap and reached fifteen common views, hoping to once again guide and advance public support for mutually beneficial integration across the Strait.

On the basis of these existing achievements, the two sides should actively promote grassroots exchanges and cooperation in economic and trade matters, culture, youth, and other fields and at all levels, continuously accumulating goodwill and deepening mutual understanding. People on both sides should travel more, make more friends, and strengthen industrial linkages so as to expand mutual benefit. By continually improving the shared well-being of both sides, we will create the strongest guarantee for the peaceful development of cross-Strait relations.

Taiwanese businesspeople and compatriots are important drivers of cross-Strait exchange and cooperation, bridges of mutual understanding between people on both sides, and the most important supporting force and supervising force for the peaceful development of cross-Strait relations. The Chinese Kuomintang has always cared about the well-being and rights protection of Taiwanese businesspeople and compatriots on the mainland. I look forward to, and believe, that in the future the mainland’s protection of the rights and interests of Taiwanese businesspeople and compatriots will certainly become even more complete.

Third, to commit ourselves to moving toward a better cross-Strait future and strengthening people’s livelihoods and well-being:

Human society is currently at the wealthiest stage in history, yet it is also an era in which development and distribution are the most unequal. The global geopolitical landscape is becoming increasingly volatile, and uncertainty in the global economy is rising accordingly. People on both sides of the Strait enjoy advanced technology and convenient lives, but the difficulties and challenges we face may also be unprecedented. I believe that no hardship or obstacle can stop the determination of people on both sides of the Strait to join hands in pursuing a better life.

Because of the closeness between the two sides in terms of geography, social patterns, cultural customs, and industrial structure, our experiences and strengths can complement each other and generate mutual benefit when dealing with modern challenges such as climate change, energy security, technology governance, and population aging. I hope the two sides can continue to strengthen exchanges and cooperation in several forward-looking fields, including energy conservation and carbon reduction, disaster prevention and mitigation, medical care and elder care, and artificial intelligence, so that together we can build a future vision that will surely help improve a community with a shared future for mankind.

The young people of both sides today are the generation in the history of the Chinese nation with the highest level of education, the broadest and deepest understanding of the world, the greatest vitality and creativity, and the clearest sense of how to use peaceful development to realize their talents. The hope of both sides lies in the youth. We should encourage young people to engage in more exchanges, explore life’s questions and development visions, and work together toward the future. When young people on both sides appreciate each other, inspire each other, and grow together, cross-Strait relations will continue to develop in a positive direction and endure over the long term.

On this trip, I take “peace across the Strait, benefiting people’s livelihoods” as my personal aspiration, and on this basis I would like to put forward the following five propositions:

First, to promote the peaceful development of cross-Strait relations:

The peaceful development of cross-Strait relations accords with the common wishes of the people on both sides and with the overall interests of the Chinese nation, and it must be firmly advanced.

Peace and development are basic human needs. People on both sides of the Strait should not stand in opposition to one another, but should live in harmony. Both sides, and both the Kuomintang and the Communist Party, have a responsibility to promote Chinese culture, to foster peace through exchange, to enhance development through cooperation, to institutionalize the peaceful development of cross-Strait relations, and to gradually achieve a peace framework.

Second, to seek the restoration of cross-Strait consultation mechanisms:

Cross-Strait consultation and contact mechanisms once played an indispensable role in peace and development for both sides, and they should be restored.

The laws and regulations on each side of the Strait both stipulate that the two sides are not state-to-state relations. In 1992, the authorized bodies of both sides reached a consensus that each side would express, by oral statement, its adherence to the one-China principle, while at the same time seeking common ground while reserving differences. This became the political foundation of cross-Strait consultation and contact mechanisms. Historical facts cannot be denied. On this basis, the consultation mechanism should be restored so as to build a virtuous cycle of goodwill.

Third, to maintain peace and stability in the Taiwan Strait and enhance mutual benefit across the Strait:

A peaceful and stable Taiwan Strait is what all parties in the region hope for, and mutually beneficial cross-Strait relations are what public opinion on both sides looks forward to. The two reinforce one another. The relevant provisions on both sides and international reality all reflect one China. On the basis of these provisions and realities, the two sides should cooperate with each other, handle differences, and consult on resolving the state of confrontation, thereby contributing to regional security. The 23 agreements on economic cooperation, including the three direct links and tariff reductions, have promoted common development and shared prosperity. Their results are clear for all to see, have been affirmed by all sectors on both sides, and should continue to be advanced on the basis of the shared political foundation so as to expand tangible benefits and strengthen public support for peace across the Strait.

Fourth, to expand Taiwan’s international participation space through political mutual trust:

Taiwan once participated in the World Health Assembly and the International Civil Aviation Organization Assembly in an appropriate manner on the basis of the 1992 Consensus, but later lost that opportunity.

In the future, once the two sides rebuild political mutual trust, they should work to enable Taiwan to return to the World Health Assembly and the International Civil Aviation Organization Assembly, and explore Taiwan’s participation in forums such as the General Assembly of INTERPOL. Regional economic integration bears directly on Taiwan’s economic development. Cross-Strait economic cooperation can be mutually reinforcing with Taiwan’s participation in regional economic integration, and the two sides should explore Taiwan’s accession to the Regional Comprehensive Economic Partnership and the Comprehensive and Progressive Agreement for Trans-Pacific Partnership.

Fifth, to continue giving full play to the communication platform between the Kuomintang and the Communist Party:

The communication platform between the Kuomintang and the Communist Party has always been a force for the right path in the peaceful development of cross-Strait relations and peace and stability in the Taiwan Strait, and it should continue to play its role.

The mechanisms of the platform, including high-level dialogue, think tank forums, youth exchanges, grassroots exchanges, and protections for Taiwanese businesspeople, have all played a leading and supportive role in opening up cross-Strait relations. At present and in the future, we should continue to use this platform to encourage exchanges, cooperation, communication, and consultation in various fields and at all levels across the Strait, so as to bring peace to the Taiwan Strait and greater well-being to the people.

Finally, I would like once again to thank the CPC Central Committee and General Secretary Xi for the invitation. Exchanges and interactions should naturally be reciprocal. I sincerely hope that one day in the future I may have the opportunity to be the host and welcome General Secretary Xi and everyone present here in Taiwan.

Thank you all.

Cheng Li-wun‘s press conference

Cheng Li-wun

First of all, I would like to express my sincere thanks to all our friends in the media for your hard work in covering this visit over the past few days. Since the day we arrived, everything has gone very smoothly. I am also deeply grateful to the Taiwan Work Office of the CPC Central Committee, as well as to our hosts in Jiangsu, Shanghai, and Beijing, for their thoughtful hospitality and the high level of reception they extended to us, which made every member of our delegation feel warmly welcomed and at ease.

Today was the centrepiece everyone had been waiting for. Early this morning, the weather was especially fine, with warm sunshine all around. In a way, it seemed to reflect the atmosphere of the meeting itself. After a gap of ten years, the leaders of our two parties met once again, and the meeting was marked by genuine feeling, candour, and sincerity. It fully demonstrated the shared aspiration, goodwill, and sincerity for peaceful development across the Strait.

This was precisely the outcome I had most sincerely hoped this visit would achieve: to send a clear message together to both sides of the Strait, to Taiwan, and to the world. Everything this morning therefore proceeded especially smoothly. About five minutes after I delivered my remarks, members of the media left the meeting room, so I would like to take this opportunity to offer a little further explanation.

In the talks that followed, I raised three main areas in which both sides can work together. The first is to preserve Chinese history and carry forward Chinese culture. The second is to enhance shared well-being and promote exchanges and cooperation. The third is to build a better cross-Strait future and improve people’s livelihoods and well-being.

During the meeting, I also put forward five proposals, guided by the aspiration for cross-Strait peace and better livelihoods. The first is to promote the peaceful development of cross-Strait relations. The second is to seek the restoration of cross-Strait consultation mechanisms. The third is to safeguard peace and stability in the Taiwan Strait and expand mutual benefit across the Strait. The fourth is to expand Taiwan’s space for international participation on the basis of political mutual trust. The fifth is to continue making good use of the communication platform between the Kuomintang and the Communist Party. We will provide our friends in the media with the full details later in a complete press release.

Today’s exchange was therefore very substantive. Over the ten years since contact was interrupted, as everyone has felt, cross-Strait relations have grown increasingly tense, and the cycle of ill will has only intensified. No one wants to see events take a negative turn. That is why I believe today’s meeting carries profound and critical significance. People on both sides of the Strait can have confidence that, so long as our starting point is sound and sincere, peaceful development across the Strait still holds every possibility for a positive and optimistic future. Today, an important first step has been taken, and what comes next will require the efforts of many more people working together.

Of course, the Chinese Kuomintang cannot shirk its responsibility. During this exchange visit, when I visited the Sun Yat-sen Mausoleum, I was especially struck by the Three Principles of the People, which are so familiar to us all: nationalism, democracy, and people’s livelihood. Yet at the mausoleum, it is people’s livelihood that is placed at the centre. When the mausoleum was built, the point that was especially emphasised was this: what are nationalism and democracy ultimately for? They are for people’s livelihood.

When we visited Shanghai, we also saw General Secretary Xi Jinping’s vision for the city: that it should be a “people’s city,” and that all of Shanghai’s development and prosperity should ultimately serve the people. In the same way, the reason we must work to resolve political division and confrontation across the Strait is also for the people, for their livelihoods, so that everyone may live a good life. The wish is really as simple and plain as that. Whatever obstacles may stand in our way, I believe that so long as our original intention is right, and so long as we persevere together, the future will surely bear fruit.

That is my brief report to you. I would now like to leave more time for your questions. Thank you.

China Times
Madam Chair, hello. This is a question from the Taiwan-based China Times. During this visit, you repeatedly mentioned the differences and divergences across the Strait. The outside world is also very concerned about whether, during the closed-door meeting this morning, you raised this issue with General Secretary Xi, especially the part about “one China, respective interpretations.”

In addition, the 1992 Consensus and the one-China principle have long been the common political foundation between the Kuomintang and the Communist Party. But it is also true that young people in Taiwan who are now in university or graduate school, as well as many first-time voters, were all born after the 1992 Consensus was reached. Most young people feel rather indifferent toward it. How will the Kuomintang show young people, or persuade party members, that the 1992 Consensus still stands the test of time and is not electoral poison? Thank you.

Cheng Li-wun
On the cross-Strait differences you mentioned at the outset, General Secretary Xi in fact addressed this point directly in the closed-door meeting just now. I took careful notes, although of course I was not able to record his remarks verbatim. Xinhua will issue a full report of the relevant content.

That in fact answers your question. He spoke specifically about the divergences you just mentioned, which I also referred to several times. He said these differences have deep historical roots, but he also stressed that we must proceed with patience and perseverance, in the spirit of Yu Gong moving mountains and Jingwei filling the sea. The freeze did not happen overnight, and it will not be resolved overnight either. But as long as there is open communication and a willingness to consult with one another, then everything can be discussed.

What struck me in particular was that General Secretary Xi said that, with regard to these cross-Strait differences, the Mainland respects the social system and way of life chosen by Taiwan compatriots, even though they differ from those on the Mainland. At the same time, he also expressed the hope that Taiwan would acknowledge the Mainland’s development achievements. This only underscores the need for more opportunities for exchange, more chances to know one another, and more opportunities to understand one another.

General Secretary Xi also stressed that meeting face to face is especially important, and that being able to see one another in person makes a very great difference. So I believe both sides share a strong common aspiration and common starting point: to narrow differences, deepen mutual understanding, expand mutual goodwill, and build greater mutual trust. These are all vital foundations for peaceful and stable cross-Strait relations in the future, and they are exactly the points I have repeatedly emphasised. As I have said, so long as something contributes to peace across the Strait, I am willing to do it; so long as a person contributes to peace across the Strait, I am willing to meet them.

General Secretary Xi also said that so long as a proposition is conducive to the peaceful development of cross-Strait relations, it should be pursued with full effort; and so long as a matter is conducive to the peaceful development of cross-Strait relations, it should likewise be pursued with full effort. This is therefore a goal and direction both sides are working toward together.

You also mentioned the 1992 Consensus and opposition to Taiwan independence. In my later remarks, I once again fully restated the substance of the 1992 Consensus. Let me think for a moment—I cannot quite remember whether General Secretary Xi said this during the meeting or over lunch—but he did specifically refer to the 1992 Consensus across the Strait. Because I had mentioned the Koo-Wang talks earlier in Shanghai, he brought that up as well, noting that those talks had in fact made the meaning of the 1992 Consensus very clear. General Secretary Xi then said that, unless one is deliberately pretending not to understand, one should not fail to see what the 1992 Consensus really means. There is therefore no need to maliciously distort it, still less to maliciously undermine reconciliation and the peaceful development of cross-Strait relations.

So, to your first point, today’s successful meeting has once again confirmed what I have been telling everyone over the past few months: that the one and only political foundation for cross-Strait exchange and dialogue is adherence to the 1992 Consensus and opposition to Taiwan independence. Of course, we must keep pace with the times by using language and forms of expression suited to the present moment, so that each generation of young people can understand the challenge we face at this stage, and understand how adherence to the 1992 Consensus and opposition to Taiwan independence can help avert war and tragedy, build peace together, and, on the basis of peace, pursue the greatest possible well-being for the people. I believe that should be the shared expectation and wish of any normal person—unless, of course, someone is intent on destroying peace, or has a special personal agenda and is willing to make the lives and property of the people of Taiwan the price of cross-Strait war. That is what we oppose, what we must stop, and what we must prevent. We hope the two sides of the Strait can be like today’s weather: calm, pleasant, and comfortable.

People’s Daily
Thank you. A question from People’s Daily. This morning, General Secretary Xi Jinping met with Chair Cheng Li-wun. What important significance do you think this has for promoting relations between the Kuomintang and the Communist Party and for the peaceful development of cross-Strait relations? Thank you.

Cheng Li-wun
I was especially grateful today that General Secretary Xi said that, on the same political foundation, the CPC Central Committee is willing to engage in exchange and dialogue not only with the Chinese Kuomintang, but with all political parties in Taiwan. I would therefore also like to say clearly that the Kuomintang and the Communist Party have a long and complicated history marked by both conflict and cooperation. The Chinese Kuomintang therefore naturally has an important responsibility to help resolve the grievances and entanglements between the two parties.

But when it comes to the peaceful development of cross-Strait relations, we also hope that all political parties in Taiwan will not treat this as a tool for party competition or vote-gathering. This issue rises above that level, because it is ultimately a choice between peace and war. I therefore very much hope that, on cross-Strait relations, all political parties in Taiwan can put aside their inter-party differences and work together for peace. General Secretary Xi also extended this significant goodwill just now. Such exchanges are absolutely not limited to the Kuomintang and the Communist Party alone. I believe this breadth of vision and openness is also something the Chinese Kuomintang very much welcomes.

We did not come here today for the private interests of one party. We came because we bear a historical responsibility, and because we cannot allow Taiwan and the Taiwan Strait to become a battlefield. So we are taking the lead. Once the road has begun to open, it will only become wider and smoother. Just as I said the other day at Yangshan Port, everyone is welcome to join; others may do even better and even more brilliantly than I have, and we would be glad to see that. So today, at a moment when the world had grown deeply pessimistic and no longer even dared to hope for anything from cross-Strait relations, the leaders of the Kuomintang and the Communist Party have shown the world that things are not as difficult as people imagine.

There was also one further point on which I felt particularly in tune with General Secretary Xi during the exchange: political leaders must not forget their original intention, and must not allow personal or partisan interests to blind them to the role political leaders ought to play and the responsibilities they ought to fulfil. I believe this is also something I have always expected of myself. At this moment, each of us is very small. But in the great arena and pivotal moments of history, one must make the right choice.

So in this process of exchange, I believe we have laid a foundation that will allow us to move ahead with greater confidence, no matter how complex and turbulent the global situation may become, and no matter what internal challenges may arise across the Strait. We must succeed; failure is not an option.

NBC
Thank you, Madam Chairwoman. When we spoke recently, you said that this trip was about seeking reconciliation with the Mainland as the best way forward for Taiwan. Having made this trip and having met with President Xi, would you now say that you share his goal of unification for Taiwan? Is that the way forward?

Cheng Li-wun
I think that, throughout today’s talks, what was truly highlighted and valued was the sense of kinship that comes from belonging to the Chinese nation. As I mentioned earlier, General Secretary Xi in fact recognised and respected Taiwan’s different way of life and system, and also hoped that this would be reciprocal—that Taiwan, too, would respect and acknowledge the Mainland’s development achievements. He also specifically said that he hopes there will be no conflict across the Strait, and that in the future both sides, as one family, can engage in more exchanges and grow closer to one another.

He also said that the freeze did not happen overnight. This requires a sustained process of effort, and it requires firm hope for the future, so that both sides may strive together in solidarity to realise the great rejuvenation of the Chinese nation. So, in this process, matters must be handled one by one, every issue addressed one by one, and the road walked step by step.

I think that, on this point, General Secretary Xi and I were both very pragmatic, and both hope to proceed step by step, just as I said earlier. At the very outset, General Secretary Xi in fact said that although social systems and political propositions may differ, our common ancestors and the bloodline of the nation cannot be severed; differences in social systems should not be used as an excuse for division.

I believe this was a very important expression of goodwill. We must face pragmatically the many differences that have arisen over the long course of cross-Strait history. But Taiwan’s achievements today and the Mainland’s achievements today are both great and remarkable achievements of the Chinese nation. We can appreciate one another, respect one another, and even learn from one another. In the future, there are even greater opportunities for cooperation, so that the achievements both sides have already attained may be expanded further, to benefit not only both sides of the Strait, but humanity as a whole.

So, in answer to your question, we hope to consolidate and strengthen a peaceful and stable relationship. On that basis, we should handle matters one by one and move forward steadily, step by step. Thank you.

China Review News
Thank you. A question from China Review News. Chair Cheng, hello. We know this is your first meeting with General Secretary Xi. At noon today, General Secretary Xi also hosted a special luncheon for you and the main members of the delegation. Could you share whether there were any details from that occasion that left a particularly deep impression on you? In addition, what important outcomes do you think this meeting achieved? Thank you.

Cheng Li-wun
I am of course very grateful for General Secretary Xi’s hospitality. We have just had a very warm luncheon. The first detail that left a particularly deep impression on me was the very first dish, because General Secretary Xi specifically said it was a Fujian dish—Fujian sea clam. This was the same dish that had been served at the state banquet when the Communist Party hosted President Nixon. He said that banquet had included two Fujian dishes, and one of them was this sea clam in chicken broth. The sea clam itself was very special. That was the first point.

The second is that General Secretary Xi was also very thoughtful and attentive to the members of our delegation. In particular, because I do not come to the Mainland very often, he asked whether I had adapted well over the past few days, whether I was in good health, and whether everything else was going smoothly. He also expressed particular concern for Chairman Lien Chan and Chairman Ma Ying-jeou, and asked us to pass on his regards and greetings to both of them. He especially recalled the Ma-Xi meeting at the time, as well as many details from Singapore. All of this gave us a very strong sense of warmth and familiarity.

Since we were also fortunate to have the Chairman of the National Development and Reform Commission present today, I specifically discussed the situation facing various sectors of Taiwanese industry, and whether there might in future be greater possibilities for alignment and cooperation. We talked about many things, including the ethnic minority group in Yunnan from which I come by birth, as well as various aspects of Taiwan’s Indigenous peoples.

So I felt that the entire conversation was very pleasant and very cordial. Thank you.

As for the achievements, I have actually already touched on them. First of all, the most important thing I hope to bring back to Taiwan is a message of peace. It is clear rather than vague, firm, and intended to continue over the long term. I believe that matters more than anything else. Beyond the issues we raised regarding Taiwanese industry, especially the situation facing traditional manufacturing and the services sector, General Secretary Xi also showed particular concern for our agricultural and fishery products. He specifically said that Taiwanese agricultural and fishery products are very welcome in the Mainland market. We also exchanged views on the expectations, conditions, and needs of different sectors in Taiwan. Thank you.

United Daily News
Chair Cheng, hello. A question from United Daily News. In your remarks, you said that you hoped to institutionalise the peaceful development of cross-Strait relations and gradually reach a framework for peace. Could you elaborate further on this concept of a peace framework? Also, in your remarks, you said you hoped to rebuild political mutual trust so that Taiwan could return to certain international organisations, and even join RCEP and the CPTPP. May I ask whether General Secretary Xi gave any concrete and positive response on these points during the closed-door meeting? Thank you.

Cheng Li-wun
In fact, General Secretary Xi’s response was especially positive. Beyond stressing our shared roots and common nationality, he said that so long as both sides achieve a meeting of minds—as I have already mentioned several times—everything can be discussed. He also stated specifically that the proposals and expectations I put forward in my remarks, every single one of them, could be actively and comprehensively studied, coordinated, and advanced. So, in our talks just now, I believe General Secretary Xi conveyed a very clear and highly positive message, and showed that he attaches great importance to Taiwan’s expectations and needs.

I should also add that he spoke specifically about Taiwanese businesspeople. Because he spent such a long time in Fujian, he knows many old friends among Taiwanese businesspeople there. He also places great importance on them. Our vice chairman quoted what General Secretary Xi had just said: with regard to the expectations we mentioned, he said he would attach great importance to them and actively consider them.

You also asked about the framework for peace. In fact, during Chairman Lien’s first Journey of Peace, many important foundations were already laid. On that basis, we have continued to work in that direction. But even more importantly, throughout this process—and General Secretary Xi referred repeatedly just now to much of the earlier history between the Kuomintang and the Communist Party, including more recent exchanges and the Xi-Ma meeting—he specifically said that we once had a very good opportunity, but unfortunately did not firmly seize it at the time. That is why, in my public remarks, I especially stressed that I hope a peaceful cross-Strait relationship can become irreversible and never move backwards.

That is why we hope for institutionalisation, and ultimately even the emergence of a cross-Strait peace framework. Once that uncertainty is removed, every possibility exists for securing a peaceful and stable future across the Strait. This requires efforts from both sides. Even more importantly, of course, we hope that the Kuomintang will return to office in 2028 and regain governing power, so that it can formally and officially represent the people of Taiwan in seeking with the Mainland the institutionalised and sustainable cross-Strait framework I have just mentioned.

We also hope that such an experience can be shared more broadly with the rest of the world, so that places where conflict may arise can draw on similar examples and experiences, turning swords into ploughshares and ensuring that war no longer takes place anywhere on earth. Thank you.

CTi News
Chair Cheng, hello. I am Chang Yang-hao from CTi News. Chair, I would like to ask: you just mentioned that, in your talks with President Xi, you raised the point that Taiwan’s space for international participation should be expanded on the basis of political mutual trust. I heard in your response just now that General Secretary Xi gave a fairly positive reply to that. Under such circumstances, does that mean these things can only be achieved after the Kuomintang comes to power? Or is it possible that, in the more immediate term, some of the county and city governments currently governed by the pan-blue/KMT camp could already carry out certain related cross-Strait exchanges?

Also, one report mentioned that, at the end of the meeting, you said you hoped that one day you might have the chance to serve as host in Taiwan and welcome everyone. Does that mean you have ambitions to move to an even higher position?

Cheng Li-wun
This is an exchange between the two parties. Today I came to the Mainland at the invitation of General Secretary Xi as the representative of the Kuomintang. So, in my capacity as Chair of the Kuomintang, I naturally also hope that, following another rotation of parties in government in the future, I may be able to invite General Secretary Xi to visit Taiwan, Penghu, Kinmen, and Matsu. So when I said that I hoped one day to serve as host in Taiwan, what I meant was simply that I genuinely look forward to, and hope for, an opportunity to invite General Secretary Xi and other leaders to Taiwan so that they may come and see it for themselves.

Phoenix TV
Chair, hello. I am Peng Shih-ting from Phoenix TV. I want to ask: since taking office, you have repeatedly stressed that you want to prove to the outside world that peace is a viable path, and one that can work. We also hope that through exchanges like this, many peace dividends can be created. But the problem is that, so long as the Kuomintang is not in power, those peace dividends are difficult to realise fully. So in the future, as the Kuomintang promotes these kinds of cross-Strait exchanges, how can it avoid having them manipulated by rival parties and still win the support of mainstream public opinion? How can that balance be struck? Thank you.

Cheng Li-wun
In fact, after I took office as party chair, all kinds of strange stories began circulating in Taiwan from various quarters. Many false accounts were fabricated, and many false messages were spread. That only exposed their unease and lack of confidence, as if they were deeply afraid that we might actually succeed in accomplishing these things, just as we have done today.

From the day I formally took office as party chair on 1 November until today, it has only been a little over four months. That is why I have repeatedly told everyone that this is not some distant and unattainable goal, nor is it something harder than ascending to heaven. I do not possess any extraordinary powers. I have stressed again and again that there are no other obstacles, no other demands, and no so-called “admission ticket” of the kind some people talk about. It comes down to one thing only: the 1992 Consensus and opposition to Taiwan independence.

Taiwan has sacrificed nothing, Taiwan has given up nothing, and yet we can already see spring returning, smiles on both sides, hands extended in greeting, and people sitting down together for exchange and dialogue. So many of the inner demons and obstacles people talk about have in fact been deliberately manufactured and manipulated.

I have also said that this will be a major electoral benefit for Taiwan. You may not all have felt it yourselves, but after ten years of interruption, with cross-Strait relations growing ever more tense and confrontational, many sectors have suffered unspeakably and have not even known whether their industries or family businesses could continue. That kind of pain and anxiety is not something politicians can brush aside with a few easy words. That is why I specifically stressed just now that all politics, in its original intention and proper breadth of mind, should take the people as its starting point.

In just these short four months, people from all walks of life and all sectors have come to see me. I have travelled all over Taiwan—I have lost count of how many times I have circled the island—and everywhere I have gone, people have consistently expressed their strong hope for peaceful exchange across the Strait, not to mention their unwillingness to see Taiwan’s next generation sent onto the battlefield. All of this will ultimately be reflected in votes.

Of course, in Taiwan, only by winning elections can we implement our political ideas and principles. That is a challenge the Kuomintang must face, and we are confronting this year’s election very seriously, with careful planning step by step. But I still want to say once again that I truly do not hope this becomes merely a calculation about elections and votes. This issue stands above that level. Yet in the face of the obstacles and opponents we may encounter, we must also overcome every difficulty and every challenge. Winning the people’s endorsement through their votes will allow the path of peace across the Strait to be pursued more steadily and more successfully.

TVBS
Chair Cheng, I am Feng Wei from TVBS. I would like to ask: just now you mentioned that you put forward five requests, and in the fourth point you spoke of expanding space for international participation. Did you raise that fourth point directly with General Secretary Xi? And was his reply positive? Secondly, there is a strong possibility that a Trump-Xi meeting will take place in May. In your remarks today, you also specifically said that you hope the Taiwan Strait will not become a chessboard for external interference. Did you exchange views with General Secretary Xi on that issue as well?

Cheng Li-wun
Regarding the fourth point, this is how I stated it: on the basis of the 1992 Consensus, Taiwan once participated, in an appropriate capacity, in the World Health Assembly and the International Civil Aviation Organization Assembly. Unfortunately, that opportunity was later lost. In the future, once political mutual trust has been rebuilt, efforts should be made to enable Taiwan to return to the World Health Assembly and the International Civil Aviation Organization Assembly, while also actively exploring Taiwan’s participation in the INTERPOL General Assembly.

In addition, regional economic integration bears directly on Taiwan’s economic development. Cross-Strait economic cooperation and Taiwan’s participation in regional economic integration can reinforce one another. Both sides may explore Taiwan’s accession to the Regional Comprehensive Economic Partnership, or RCEP, and the Comprehensive and Progressive Agreement for Trans-Pacific Partnership, or CPTPP. That was the substance of my remarks just now.

Taken as a whole, with regard to these and other requests and proposals we raised, as I said, General Secretary Xi viewed them and responded to them all very positively.

TVBS
And the Trump-Xi meeting?

Cheng Li-wun
No, that was not mentioned. That was not brought up.

Unlike previous fragmented tools, this 18-article regulation represents the first comprehensive administrative rule in this space. It authorizes Chinese authorities to review measures taken by foreign governments, institutions, or individuals—particularly those involving discriminatory restrictions or disruptions affecting China’s industrial ecosystem—and to respond with corresponding actions.

The rules also lay out a menu of potential responses, ranging from trade and investment curbs to additional charges and restrictions on market access. Overall coordination will sit with the State Council, while implementation will be carried out by relevant agencies across foreign affairs, security, and legal domains.

According to the interpretation of China’s Ministry of Justice, the new framework is intended to close a long-standing gap in China’s legal toolkit and enhance the country’s ability to manage risks to supply-chain stability. It also reorganizes existing policy instruments into a more structured system that integrates investigation procedures, risk monitoring, early-warning signals, and emergency response mechanisms—while maintaining a broader commitment to international engagement.

Ministry of Justice Official Answers Questions on the Regulation on Industrial and Supply Chain Security

Source: Ministry of Justice | April 7, 2026

On March 31, 2026, Premier Li Qiang signed State Council Decree No. 834, issuing the Regulation on Industrial and Supply Chain Security (the “Regulation”), which took effect immediately upon release. A senior official from the Ministry of Justice responded to questions from the press.

Q: Could you briefly explain the background of the Regulation?

A: The CPC Central Committee and the State Council attach great importance to industrial and supply chain security. Xi Jinping has repeatedly emphasized the need to advance both technological and institutional innovation, address bottlenecks and vulnerabilities in supply systems, and enhance both competitiveness and security.

Recent Party plenums have also called for improving mechanisms to strengthen supply chain resilience. At the same time, a number of countries have introduced legislation and national strategies to safeguard their own supply chains.

While China already has rules in areas such as emergency response, export controls, and countermeasures, there has been no dedicated legal framework specifically for supply chain security. This Regulation is intended to fill that gap, strengthen the legal foundation, and ensure China can better respond to both domestic and external risks while maintaining stable and efficient supply chains.

Q: What is the overall approach behind the Regulation?

A: There are three main principles:

First, balancing development and security. The Regulation promotes high-quality development of industrial and supply chains, while also establishing mechanisms to manage risks.

Second, focusing on urgent priorities. It targets key sectors related to economic stability and national security, using a “small but focused” approach and ensuring alignment with existing policies.

Third, being problem-oriented. It addresses real-world challenges by establishing mechanisms for risk monitoring, early warning, and emergency response, while also introducing investigation procedures, countermeasures, and extraterritorial provisions.

Q: What are the key principles governing supply chain security work?

A: The Regulation reflects several core principles:

It follows a holistic national security approach, balancing development and security, domestic and international priorities, while promoting high-level opening-up and stable global supply chains.

It supports R&D in key technologies, strengthens core capabilities, and promotes digitalization and intelligent upgrading of supply chains.

It adheres to principles of mutual benefit and win-win cooperation, while actively participating in the development of international rules.

Q: How does the Regulation prioritize key areas?

A: Given the complexity of supply chains, the Regulation adopts a focused approach by establishing a “key sectors list.” This allows authorities to concentrate resources and coordination on areas critical to economic and national security.

For these sectors, the Regulation sets out clear requirements:

Strengthening data sharing while ensuring data security

Building risk monitoring and early-warning systems

Enhancing risk prevention through stockpiling and capability building

Establishing emergency response mechanisms, including contingency plans and rapid resource mobilization

Q: How does the Regulation address harmful foreign actions?

A: Articles 14 and 15 define two main scenarios:

First, if foreign governments, regions, or international organizations adopt discriminatory restrictions or similar measures in violation of international norms—and harm or support actions that harm China’s supply chains—relevant authorities may launch investigations and take countermeasures. These may include restricting trade, imposing fees, or adding entities to countermeasure lists under laws such as the Anti-Foreign Sanctions Law.

Second, if foreign companies or individuals violate normal market practices—such as disrupting transactions with Chinese entities or applying discriminatory treatment—authorities may also initiate investigations. Depending on the findings, measures may include restricting trade and investment, banning cooperation, limiting entry, or revoking residency and work eligibility. These measures can also apply to entities controlled or operated by the parties involved.

Q: How does the Regulation ensure information security?

A: The Regulation establishes a framework for supply-chain-related data security:

Unauthorized data collection related to supply chains within China is prohibited, and violations will be handled according to law.

Companies and research institutions are required to strengthen risk management systems to ensure control over core technologies, systems, and data, with government support through guidance and training.

Enterprises and industry associations are encouraged to report risks or incidents affecting supply chain security to relevant authorities.

In practical terms, authorities may launch security reviews under two broad conditions: when external actors introduce measures that violate international norms and discriminate against China, or when foreign firms or individuals disrupt normal commercial interactions with Chinese entities in a way that causes, or could cause, significant harm. Confirmed violations could lead to restrictions on trade, investment, or cooperation, as well as entry bans for relevant individuals.

The regulation places particular emphasis on sectors deemed vital to economic and national security. Authorities will define and regularly update these priority areas, while building a “closed-loop” governance system covering data sharing, risk tracking, early warning, and crisis management. It also explicitly prohibits unauthorized collection of supply-chain-related data within China and provides legal backing for enforcement actions.

Beyond the regulatory details, several Chinese experts interpret the move as part of a broader strategic recalibration.

Wang Minghui from the State Council’s Development Research Center argues that industrial and supply chains function as the backbone of a major economy. In his view, the new rules are a key step in responding to a more uncertain global environment, addressing technological choke points, and strengthening the legal foundation for industrial security. More fundamentally, he frames the policy as advancing a dual objective: using security to support development, and using development to reinforce security—helping China transition toward a more robust and resilient industrial structure.

Li Jin, a specialist in state-owned enterprise reform, notes that while China’s industrial base is broadly stable, weaknesses remain in higher-end segments. He sees the regulation as a targeted effort to reinforce these weak links and as a signal that China is moving into a new phase of institutionalized supply-chain governance, with implications for long-term growth quality.

From a legal standpoint, Liao Shiping of Beijing Normal University highlights that a major innovation lies in making countermeasures more specific and actionable. By linking with existing frameworks such as the Foreign Relations Law and the Anti-Foreign Sanctions Law, the regulation contributes to building a more coherent system of foreign-related rule of law. He emphasizes that the approach balances openness and security: externally, China continues to support cooperation rather than decoupling; internally, it strengthens its ability to defend against risks and respond when necessary.

Yang Daqing of the China Federation of Logistics and Purchasing underscores that the policy is not about retreating into self-sufficiency, but about enabling a higher level of openness on a more secure footing. He points to the importance of leading firms anchoring supply chains, stronger coordination across upstream and downstream players, and a combination of domestic backup capacity and diversified global landscape to enhance resilience and stability.

Taken together, the regulation not only expands China’s policy toolkit but also reflects a broader shift toward proactive, rules-based management of supply-chain risks. It calls for closer coordination across agencies, improved information sharing, more sophisticated early-warning systems, and better preparedness for potential disruptions—while signaling a willingness to respond more assertively to external pressure when necessary.

Full translation (unofficial) of the regulation:

Regulation of the State Council on Industrial and Supply Chain Security

Article 1
This Regulation is formulated in accordance with the National Security Law of the People’s Republic of China, the Foreign Relations Law of the People’s Republic of China, the Anti-Foreign Sanctions Law of the People’s Republic of China, the Foreign Trade Law of the People’s Republic of China, and other relevant laws, for the purposes of preventing risks to industrial and supply chain security, enhancing resilience and security levels, and safeguarding economic and social stability as well as national security.

Article 2
Work on industrial and supply chain security shall follow a holistic national security approach, coordinate development and security, balance domestic and international considerations, promote high-level opening-up, and facilitate the stable and smooth operation of global industrial and supply chains.

Article 3
The State shall establish and improve a working mechanism for industrial and supply chain security, and coordinate related efforts in a unified manner.

Relevant departments under the State Council—including those responsible for foreign affairs, development and reform, industry and information technology, public security, national security, rule of law, finance, natural resources, transport, agriculture and rural affairs, commerce, financial regulation, customs, market regulation, and cyberspace administration—shall, according to their respective responsibilities, undertake work related to industrial and supply chain security and strengthen coordination and cooperation.

People’s governments of provinces, autonomous regions, and municipalities directly under the central government shall, under national coordination, be responsible for industrial and supply chain security work within their respective administrative regions.

Article 4
The State shall guide the rational and orderly layout of industrial and supply chains, promote their digitalization and intelligent transformation, enhance their security and controllability, and advance high-quality development.

Enterprises are encouraged and supported to diversify supply channels, strengthen industrial and supply chain cooperation, participate fairly in market competition, and improve their capacity to prevent related risks.

Article 5
The State shall adhere to the principles of equality, mutual benefit, and win-win cooperation, strengthen international cooperation in the field of industrial and supply chains, and actively participate in the formulation of relevant international rules.

Article 6
Relevant departments of the State Council and local governments shall, when formulating relevant plans, fully consider potential impacts on industrial and supply chain security.

Article 7
The State shall strengthen security guarantees for industrial and supply chains in key sectors. Relevant departments under the State Council shall formulate and dynamically adjust a list of key sectors to ensure stable and continuous production and circulation of raw materials, technologies, equipment, and products in these sectors.

Article 8
Relevant departments of the State Council shall promote information sharing on industrial and supply chains in key sectors, strengthen platform support, guide industries and enterprises to enhance connectivity, and adopt effective measures to ensure data security.

Article 9
The State shall establish and improve a monitoring and early-warning system for industrial and supply chain security risks in key sectors. Relevant departments shall organize assessments and monitoring of the stability of supply channels for raw materials, technologies, equipment, and products, and their impact on economic and social stability and national security, identify risks, and promptly issue early-warning information.

Enterprises, industry associations, and chambers of commerce that identify situations affecting supply chain security may report them to relevant government departments at or above the county level.

Article 10
The State shall establish and improve a risk prevention system for industrial and supply chain security in key sectors. Relevant departments shall organize physical reserves and capacity reserves, increase efforts in R&D of technologies, equipment, and products, and enhance resilience.

Departments under the State Council and local governments shall adopt targeted risk prevention measures based on industry and regional characteristics.

Article 11
The State shall establish and improve an emergency management system for industrial and supply chain security in key sectors. Relevant departments shall formulate emergency response plans. Where situations arise that affect supply chain security and endanger economic and social stability or national security, emergency measures—such as resource mobilization, use of reserves, and organization of production, transportation, and supply—may be taken upon decision by the State Council or authorized departments. Where laws or administrative regulations provide otherwise, such provisions shall apply.

Relevant departments and local governments shall implement emergency measures according to their responsibilities and terminate them promptly once the situation is resolved. Relevant organizations and individuals shall cooperate with emergency measures.

Article 12
The State encourages and guides social capital investment to support scientific and technological R&D and breakthroughs in core technologies in key sectors, and to promote the commercialization of technological outcome.

Enterprises and research institutions shall improve risk prevention systems to ensure the security and controllability of core technologies, information systems, and data. Relevant departments shall provide guidance and training.

Article 13
Where any organization or individual conducts investigations or collects information related to industrial and supply chains within China in violation of laws, administrative regulations, departmental rules, or relevant state provisions, competent authorities shall take appropriate measures in accordance with the law.

Article 14
Where foreign states, regions, or international organizations, in violation of international law or basic norms of international relations, adopt discriminatory prohibitions, restrictions, or similar measures against China in the field of industrial and supply chains, and implement or assist in implementing acts that harm China’s supply chain security, relevant departments of the State Council have the authority to initiate investigations.

According to procedures, corresponding measures may be taken, including but not limited to restricting or prohibiting imports and exports of goods and technologies, international trade in services, or imposing special fees.

Relevant departments may, in accordance with the Anti-Foreign Sanctions Law and its implementing provisions, place organizations or individuals directly or indirectly involved in such actions on countermeasure lists and take corresponding countermeasures.

Article 15
Where foreign organizations or individuals violate normal market transaction principles, disrupt normal transactions with Chinese citizens or organizations, adopt discriminatory measures, or engage in other conduct that causes or threatens substantial harm to China’s supply chain security, relevant departments of the State Council have the authority to initiate investigations.

Investigations may include inquiries, review or copying of relevant documents and materials, and other necessary measures. Relevant parties shall cooperate and may present statements and defenses during the investigation.

Based on investigation results, authorities may impose measures such as restricting or prohibiting import and export activities, limiting investment in China, restricting cooperation with Chinese entities, denying entry of relevant individuals or transport, or revoking work, stay, or residence qualifications. Such measures may also apply to entities controlled or operated by the parties concerned.

Article 16
Organizations and individuals within China shall comply with measures adopted by relevant State Council departments under Articles 14 and 15.

For violations, authorities may order corrections and impose measures such as restricting participation in government procurement or bidding, limiting trade and service activities, restricting cross-border data transfers, or restricting exit or residence.

Article 17
Law firms, notary offices, and other professional service institutions are encouraged and supported to provide legal services related to industrial and supply chain security.

Article 18
This Regulation shall enter into force on the date of promulgation.

Since the Opinions on Strengthening the Governance of Science and Technology Ethics in 2022 established the top-level design, China has gradually elevated artificial intelligence—alongside the life sciences—as a key domain of governance. The Measures for the Ethical Review of Science and Technology (Trial) released in 2023, together with subsequent regulations on generative AI, established a core mechanism centered on “primary responsibility for self-review by institutions, supplemented by expert review for high-risk cases,” while tightly linking ethical compliance with algorithm filing and security assessments. In particular, AI technologies with capabilities for public opinion mobilization, highly autonomous decision-making, or deep human–machine integration have been explicitly included in the “high-risk science and technology activities list,” requiring mandatory review by third-party experts.

The introduction of the Administrative Measures for the Ethical Review and Services of Artificial Intelligence Science and Technology (Trial) marks a new stage in China’s AI ethics governance, characterized by a dual emphasis on professionalization and service provision.

This system not only requires companies to submit proof of ethical review during algorithm filing, but also introduces, for the first time, a dual-track access model combining “algorithm filing + ethical evaluation.” At the same time, regulatory attention is expanding beyond content security to encompass broader societal and labor protections. For example, “algorithm auditing” mechanisms targeting platform-based sectors such as ride-hailing and food delivery now mandate that algorithmic systems incorporate human override functions to prevent “algorithmic exploitation” of workers. China is thus transforming AI governance from a focus on “content and security regulation” into a more institutionalized, operational, and auditable “ethical compliance system,” embedded within the broader framework of national technology governance and industrial policy.

In terms of institutional design, the system rests on a three-tier structure: internal ethics committees within organizations, external service centers, and government-led expert review. All universities, research institutions, and companies engaged in AI development are required to establish ethics committees and assume primary responsibility. Where internal capacity is insufficient, organizations may delegate to external “ethics review service centers.” For high-risk projects—such as those affecting public opinion, psychological behavior, or involving highly automated decision-making—mandatory entry into a government-led expert review process is required. In essence, this design embeds AI ethics governance within organizations while preserving the state’s ultimate supervisory authority.

At the operational level, the Measures establish a relatively comprehensive quasi-administrative approval process. Before a project can commence, applicants must submit detailed materials, including technical plans, data sources, algorithmic mechanisms, application scenarios, and, most critically, ethical risk assessments and contingency plans. Review bodies are required to issue decisions within 30 days and may request revisions or reject applications outright. Approval is not a one-off process: projects are subject to ongoing monitoring and review during operation, and may be suspended or terminated if risk conditions change. This implies that AI projects in China will be subject to a form of “dynamic regulation” similar to that applied to pharmaceuticals or medical research.

From the perspective of review criteria, the Measures effectively establish an operationalized indicator system for AI governance in China. Six key dimensions are emphasized: whether the technology genuinely promotes social well-being; whether it entails algorithmic discrimination; whether the system is controllable and reliable; whether it is transparent and explainable; whether accountability is traceable; and whether privacy is adequately protected. While these standards appear broadly aligned with Western AI governance frameworks—such as the OECD principles and the EU AI Act—their implementation places greater emphasis on controllability and risk prevention, reflecting a more engineering-oriented approach to governance.

The most distinctive feature of this framework is that it is not a traditional regulatory model focused solely on review without support; rather, it embeds a systematic service provision mechanism alongside oversight. On the one hand, it establishes compliance boundaries through tools such as ethical review and expert reassessment; on the other, it provides enterprises with risk identification and compliance capabilities through ethics review service centers and standardized evaluation tools. In essence, it transforms AI ethics from a mere compliance threshold into a capability that can be provided and outsourced, reflecting a governance approach that both manages risk and promotes development.

Full translation of the Measures (Unofficial):

Interim Measures for the Ethical Review and Services of Artificial Intelligence Science and Technology Activities

Chapter I General Provisions

Article 1
In order to regulate the ethical governance of artificial intelligence science and technology activities, promote fairness, justice, harmony, safety, and responsible innovation, and facilitate the healthy development of the artificial intelligence industry, these Measures are formulated in accordance with the Law of the People’s Republic of China on Scientific and Technological Progress, the Opinions on Strengthening the Governance of Science and Technology Ethics, the Measures for the Ethical Review of Science and Technology (Interim) (hereinafter referred to as the “Ethics Measures”), and other relevant laws, regulations, and provisions.

Article 2
The artificial intelligence science and technology activities to which these Measures apply refer to artificial intelligence scientific research, technological development, and other activities conducted within the territory of the People’s Republic of China that may pose ethical risks or challenges in terms of human dignity, public order, life and health, ecological environment, and sustainable development, as well as other science and technology activities that are required to undergo artificial intelligence science and technology ethics review in accordance with laws, administrative regulations, and relevant national provisions.

Article 3
Entities conducting artificial intelligence science and technology activities shall integrate ethical requirements throughout the entire process, adhere to the principles of promoting human well-being, respecting life and rights, ensuring fairness and justice, reasonably controlling risks, maintaining openness and transparency, protecting privacy and security, and ensuring controllable and trustworthy artificial intelligence, and shall comply with the Constitution, laws and regulations of China, and relevant provisions.

Chapter II Services and Promotion

Article 4
A system of artificial intelligence science and technology ethics standards shall be established and improved, and efforts shall be made to promote the formulation of relevant international standards, national standards, industry standards, and group standards, and to support the establishment of platforms for international standardization exchanges and cooperation.
Higher education institutions, research institutions, medical and health institutions, enterprises, and scientific and technological social organizations are encouraged to participate in the formulation, validation, and promotion of artificial intelligence science and technology ethics standards.

Article 5
The construction of an artificial intelligence science and technology ethics service system shall be advanced, strengthening the supply of services such as artificial intelligence science and technology ethics risk monitoring and early warning, testing and evaluation, certification, and consulting, improving enterprises’ capabilities in technological research and development and artificial intelligence science and technology ethics risk prevention, increasing support and service efforts for small, medium, and micro enterprises in artificial intelligence science and technology ethics review, and promoting international exchange and cooperation in artificial intelligence science and technology ethics.

Article 6
Higher education institutions, research institutions, medical and health institutions, enterprises, and scientific and technological social organizations are encouraged to carry out research on artificial intelligence science and technology ethics review, support technological innovation in artificial intelligence science and technology ethics review, strengthen the use of technical means to prevent artificial intelligence science and technology ethics risks; promote the orderly open sharing of high-quality datasets for artificial intelligence science and technology ethics review, strengthen the research and development of general risk management, evaluation, and auditing tools, explore science and technology ethics risk assessment and evaluation based on application scenarios; promote artificial intelligence products and services that comply with science and technology ethics, and protect intellectual property rights related to artificial intelligence science and technology ethics review technologies.

Article 7
Publicity and education on artificial intelligence science and technology ethics shall be carried out, the role of scientific and technological social organizations in artificial intelligence science and technology ethics publicity and education shall be brought into play, public participation shall be encouraged, practical demonstrations shall be promoted, and public awareness and literacy in ethics shall be enhanced. Mass media shall be guided to conduct targeted publicity and education on artificial intelligence science and technology ethics.

Article 8
Support shall be provided to higher education institutions, research institutions, medical and health institutions, enterprises, and scientific and technological social organizations to carry out education and training related to artificial intelligence science and technology ethics, promote the development of professional systems and curriculum systems, cultivate artificial intelligence science and technology ethics talents through multiple approaches, and promote talent exchange.

Chapter III Implementing Entities

Article 9
Higher education institutions, research institutions, medical and health institutions, enterprises, and other entities engaged in artificial intelligence science and technology activities shall be the responsible entities for the management of artificial intelligence science and technology ethics review within their organizations, and shall establish artificial intelligence science and technology ethics committees (hereinafter referred to as the “Committee”) in accordance with the relevant requirements of Article 4 of the Ethics Measures.
The Committee shall be equipped with necessary personnel, office premises, funding, and other conditions, and effective measures shall be taken to ensure that the Committee can independently carry out its work. Qualified relevant entities are encouraged to carry out certification related to artificial intelligence science and technology ethics management systems.

Article 10
The charter, composition, and the responsibilities and obligations of Committee members shall comply with Articles 5 to 8 of the Ethics Measures. The composition of the Committee shall include experts with corresponding professional backgrounds in artificial intelligence technology, applications, ethics, law, and other fields.

Article 11
Local authorities and relevant competent departments may, in light of actual circumstances, rely on relevant entities to establish specialized artificial intelligence science and technology ethics review and service centers (hereinafter referred to as “Service Centers”).
Service Centers may accept commissions from other entities to provide services such as artificial intelligence science and technology ethics review, re-examination, training, and consulting. A Service Center shall not simultaneously provide both review and re-examination services for the same artificial intelligence science and technology activity.
Service Centers shall establish standardized management systems and procedures, be equipped with full-time personnel with the capability to conduct artificial intelligence science and technology ethics review and services, and shall be subject to supervision by local or relevant competent departments.

Chapter IV Working Procedures

Section I Application and Acceptance

Article 12
For artificial intelligence science and technology activities falling within the scope specified in Article 2 of these Measures, the person in charge of the artificial intelligence science and technology activity shall apply to the Committee of their entity. Where the entity has not established a Committee or the Committee is unable to meet the requirements for conducting science and technology ethics review, an application shall be submitted to the Service Center entrusted by the entity to conduct the ethics review; where there is no affiliated entity, a qualified Service Center shall be entrusted to conduct the ethics review.
The person in charge of the artificial intelligence science and technology activity shall, in accordance with the provisions, submit application materials to the Committee or the Service Center. The application materials shall mainly include:
(1) the artificial intelligence science and technology activity plan, including research background, objectives and plans, legal qualification materials of the relevant institutions involved, personnel information, sources of funding, the algorithm mechanisms and principles to be adopted, data sources and methods of acquisition, testing and evaluation methods, the software and hardware products to be formed, expected application fields and applicable groups, etc.;
(2) the ethical risk assessment of the artificial intelligence science and technology activity, as well as prevention, control, and emergency response plans, including assessment of potential ethical risks arising from the expected application of artificial intelligence technology, monitoring and early warning measures for ethical risks, and prevention and control plans for potential ethical risks;
(3) a letter of commitment to comply with artificial intelligence science and technology ethics and research integrity requirements.

Article 13
The Committee or the Service Center shall determine whether to accept the application based on the submitted materials and notify the applicant. Where the application is accepted, the applicable procedure—general, simplified, or emergency—shall be determined based on factors such as the likelihood and degree of occurrence of ethical risks and emergency circumstances. Ethics review shall be conducted through offline or online forms as required by different procedures. Where the materials are incomplete, the applicant shall be informed in a one-time comprehensive manner of the materials that need to be supplemented.

Section II General Procedures and Simplified Procedures

Article 14
Meetings for artificial intelligence science and technology ethics review shall be chaired by the Chairperson of the Committee or a Vice-Chairperson designated by the Chairperson. No fewer than five members shall be present, and members from different categories as specified in Article 10 of these Measures shall be included. Service Centers may organize and implement their work with reference to the Committee’s provisions.
Based on review needs, experts or consultants in relevant fields who do not have a direct interest in the matter may be invited to provide advisory opinions. Advisory experts shall not participate in voting.

Article 15
When conducting artificial intelligence science and technology ethics review, the Committee or the Service Center shall focus on the following aspects:
(1) In terms of human well-being, whether the artificial intelligence science and technology activity has scientific and social value; whether the research objectives contribute positively to enhancing human well-being and achieving sustainable social development; and whether the risks of the activity are reasonably balanced against its benefits.
(2) In terms of fairness and justice, whether the standards for selecting training data and the design of algorithms, models, and systems are reasonable; whether measures have been taken to prevent bias, discrimination, and algorithmic exploitation, and to ensure objectivity and inclusiveness in resource allocation, access to opportunities, and decision-making processes.
(3) In terms of controllability and trustworthiness, whether the robustness of models and systems can be ensured to cope with open environments, extreme situations, and interfering factors; whether users are able to control, guide, and intervene in the basic operation of models and systems; and whether continuous monitoring plans and emergency response plans have been formulated.
(4) In terms of transparency and explainability, whether information such as the purpose, operational logic, interaction methods, and potential risks of algorithms, models, and systems is reasonably disclosed; and whether effective technical means are adopted to enhance explainability.
(5) In terms of accountability and traceability, whether measures such as log management are in place to clearly record sufficient information on data, algorithms, models, and systems at each stage, ensuring full-chain traceability and management; and whether the qualifications of scientific and technical personnel meet relevant requirements.
(6) In terms of privacy protection, whether sufficient measures are taken to ensure the effective protection of privacy data in activities such as data collection, storage, processing, and use, as well as in the research and development of new data technologies.

Article 16
The Committee or the Service Center shall, within 30 days from the date of acceptance of the application, make a decision of approval, approval after modification, or disapproval. In complex cases or where supplementary or corrective materials are required, the time limit may be appropriately extended, and the extended period shall be specified.
For cases requiring modification or disapproval, the Committee or the Service Center shall provide suggestions for modification or state the reasons. Where the applicant has objections, they shall file an appeal with the Committee or the Service Center within three working days from the date of receipt of the decision. Where the grounds for appeal are sufficient, the Committee or the Service Center shall make a new decision within seven working days.

Article 17
The person in charge of the artificial intelligence science and technology activity shall promptly identify changes in ethical risks and report such changes to the Committee or the Service Center.
The Committee or the Service Center shall, in accordance with Article 19 of the Ethics Measures, conduct follow-up reviews of approved artificial intelligence science and technology activities, promptly grasp changes in ethical risks, and may, where necessary, make decisions such as suspending or terminating relevant activities. The interval for follow-up reviews shall generally not exceed 12 months.

Article 18
Where multiple entities jointly carry out artificial intelligence science and technology activities, mutual recognition of ethics review results among entities may be conducted based on actual circumstances.

Article 19
A simplified procedure may be applied under any of the following circumstances:
(1) the likelihood and degree of ethical risks of the artificial intelligence science and technology activity are not higher than the routine risks encountered in daily life;
(2) minor modifications are made to an already approved activity plan without increasing the risk-benefit ratio;
(3) follow-up reviews of activities without major adjustments in earlier stages.

Article 20
The Committee or the Service Center shall formulate working procedures and tracking frequency for reviews under the simplified procedure. Simplified reviews shall be conducted by two or more members designated by the Chairperson of the Committee. Service Centers may organize implementation with reference to the Committee’s provisions.
Where, during the simplified review process, a negative opinion arises, doubts exist regarding the review content, or members’ opinions are inconsistent, the case shall be transferred to the general procedure.

Section III Expert Re-examination Procedures

Article 21
The Ministry of Industry and Information Technology and the Ministry of Science and Technology, together with relevant departments, shall formulate and publish a “List of Artificial Intelligence Science and Technology Activities Requiring Expert Re-examination” (hereinafter referred to as the “Re-examination List”), and dynamically adjust it as needed.

Article 22
For artificial intelligence science and technology activities included in the Re-examination List, after passing the preliminary review by the Committee or the Service Center, an application for expert re-examination shall be submitted by the entity. Where multiple entities are involved, the leading entity shall be responsible for the application.
Central enterprises, as well as higher education institutions, research institutions, and medical and health institutions directly under central and state organs, shall directly submit applications to the relevant competent departments for organizing expert re-examination. Other entities shall submit applications to local authorities for organizing expert re-examination.

Article 23
The entity undertaking the artificial intelligence science and technology activity shall submit materials for expert re-examination in accordance with Article 27 of the Ethics Measures.
Local or relevant competent departments shall, in accordance with Articles 28 to 30 of the Ethics Measures, organize the establishment of expert review groups to review the compliance and rationality of the preliminary review opinions, and shall provide feedback on the re-examination opinions to the applying entity within 30 days of receiving the application.
Local or relevant competent departments may entrust Service Centers to carry out specific re-examination work.

Article 24
The Committee or the Service Center shall make a final ethics review decision based on the expert re-examination opinions.

Article 25
The Committee or the Service Center shall strengthen follow-up reviews of artificial intelligence science and technology activities included in the Re-examination List, with intervals generally not exceeding six months.
Where there are significant changes in ethical risks, a new ethics review shall be conducted in accordance with Article 20 of the Ethics Measures and an application for expert re-examination shall be submitted.

Article 26
Where artificial intelligence science and technology activities are subject to regulatory measures such as registration, filing, or administrative approval in areas including deep synthesis, algorithmic recommendation, and generative artificial intelligence service management, and where compliance with ethical requirements is incorporated as a condition for approval or regulatory content, expert re-examination may no longer be required.

Section IV Emergency Procedures

Article 27
The Committee or the Service Center shall establish emergency review systems for artificial intelligence science and technology ethics, specifying emergency review processes and standard operating procedures under urgent circumstances such as public emergencies. Emergency reviews shall generally be completed within 72 hours. For activities subject to expert re-examination procedures, the review prior to expert re-examination shall generally be completed within 36 hours.

Article 28
The Committee or the Service Center shall ensure the quality and timeliness of emergency ethics reviews, strengthen follow-up work and process supervision, and, where necessary, may invite advisory experts in relevant fields to attend meetings and provide opinions.

Chapter V Supervision and Administration

Article 29
The Ministry of Science and Technology shall be responsible for overall coordination and guidance of national science and technology ethics supervision. The Ministry of Industry and Information Technology, together with relevant departments, shall be responsible for artificial intelligence science and technology ethics governance and strengthen coordination and guidance of emergency ethics reviews.
Relevant departments shall, within the scope of their respective responsibilities, supervise and administer artificial intelligence science and technology ethics reviews within their industries and systems. Local authorities shall, within the scope of their responsibilities, supervise and administer artificial intelligence science and technology ethics reviews within their jurisdictions.

Article 30
Entities shall, in accordance with Articles 43 to 45 of the Ethics Measures, register relevant information of Committees and artificial intelligence science and technology activities included in the Re-examination List through the National Science and Technology Ethics Management Information Registration Platform, and submit annual reports on Committee work and implementation reports of activities included in the Re-examination List, among other materials.
Service Centers shall register and submit annual work reports in accordance with the above provisions.
The Ministry of Science and Technology and relevant competent departments shall share information related to artificial intelligence science and technology ethics registration.

Article 31
Local authorities, relevant competent departments, and entities engaged in artificial intelligence science and technology activities shall, based on the actual conditions of their industries, systems, and entities, establish smooth channels for reporting violations of artificial intelligence science and technology ethics, and handle such matters in accordance with relevant provisions.

Article 32
Where violations of these Measures occur in the course of artificial intelligence science and technology activities or related ethical work, they shall be investigated and handled, and corresponding penalties imposed, in accordance with the Cybersecurity Law of the People’s Republic of China, the Data Security Law of the People’s Republic of China, the Personal Information Protection Law of the People’s Republic of China, the Law of the People’s Republic of China on Scientific and Technological Progress, and other relevant laws, regulations, and provisions.

Chapter VI Supplementary Provisions

Article 33
For time limits stipulated in these Measures, where not specified as working days, they shall be counted as calendar days.
The term “local” as used in these Measures refers to provincial-level administrative departments designated by provincial people’s governments to be responsible for the ethics review and management of artificial intelligence science and technology. The term “relevant competent departments” refers to relevant departments under the State Council.

Article 34
Local authorities and relevant competent departments may, in accordance with these Measures and based on actual circumstances, formulate or revise rules and detailed measures for artificial intelligence science and technology ethics review and services within their respective regions, industries, or systems. Scientific and technological social organizations may formulate specific norms and guidelines for ethics review and services in their respective fields.

Article 35
Where relevant competent departments have special provisions for artificial intelligence science and technology ethics review and services within their industries or systems that are consistent with the spirit of these Measures, such provisions shall prevail. Matters not provided for in these Measures shall be governed by the Ethics Measures and relevant laws and regulations.

Article 36
These Measures shall be interpreted by the Ministry of Industry and Information Technology in conjunction with relevant departments.

Article 37
These Measures shall come into force upon the date of issuance.

Annex: List of Artificial Intelligence Science and Technology Activities Requiring Expert Re-examination

1. The development of human–machine integration systems that have a significant impact on human behavior, psychological emotions, and life and health.

2. The development of algorithm models, application programs, and systems that possess the capability to influence public opinion, social mobilization, and social consciousness.

3. The development of highly autonomous automated decision-making systems for scenarios involving safety risks and risks to personal health.

This list shall be dynamically adjusted as required.

The first investigation targets “U.S. practices and measures that disrupt global production and supply chains.” The Chinese government believes that the United States has implemented a large number of practices and measures in trade-related fields that seriously disrupt global production and supply chains, including but not limited to: restricting or prohibiting Chinese products from entering the U.S. market, restricting or prohibiting the export of high-tech products to China, and restricting or prohibiting two-way investment in key areas. The above practices and measures may seriously harm the trade interests of Chinese enterprises, and some of them are suspected of violating WTO rules and other economic and trade treaties or agreements jointly concluded or acceded to by China and the United States.

Ministry of Commerce Announcement No. 17 of 2026
Initiation of Trade Barrier Investigation on U.S. Practices and Measures Disrupting Global Production and Supply Chains

Issuing Authority: Bureau of Trade Remedy Investigations
Announcement No.: Ministry of Commerce Announcement No. 17 of 2026
Date of Issuance: March 27, 2026

In accordance with the relevant provisions of the Foreign Trade Law of the People’s Republic of China and the Rules on Foreign Trade Barrier Investigations of the Ministry of Commerce, the Ministry of Commerce may, on its own initiative, conduct investigations into trade barriers of relevant countries and regions in order to safeguard the order of foreign trade.

Preliminary evidence and information obtained by the Ministry of Commerce show that the United States has implemented a large number of practices and measures in trade-related fields that seriously disrupt global production and supply chains, including but not limited to: restricting or prohibiting Chinese products from entering the U.S. market, restricting or prohibiting the export of high-tech products to China, and restricting or prohibiting two-way investment in key areas. These practices and measures may seriously harm the trade interests of Chinese enterprises, and some of them are suspected of violating WTO rules and other economic and trade treaties or agreements jointly concluded or acceded to by China and the United States.

In accordance with Articles 41 and 42 of the Foreign Trade Law of the People’s Republic of China and Articles 12 and 35 of the Rules on Foreign Trade Barrier Investigations, the Ministry of Commerce has decided to initiate a trade barrier investigation into the relevant U.S. practices and measures as of March 27, 2026. The relevant matters are hereby announced as follows:

I. Investigated Measures
The investigated measures in this case are: the practices and measures implemented by the United States in trade-related fields that disrupt global production and supply chains.

II. Investigation Procedures
In accordance with the Rules on Foreign Trade Barrier Investigations, the Ministry of Commerce may conduct the investigation by means of questionnaires, hearings, on-site investigations, and other methods to gather information from interested parties.

III. Investigation Period
This case shall be concluded within 6 months from the date of the announcement of the decision to initiate the investigation. Under special circumstances, the period may be extended, but the extension shall not exceed 3 months.

IV. Access to Public Information
Interested parties may download the public information of this case from the sub-website of the Bureau of Trade Remedy Investigations on the Ministry of Commerce website, or go to the Ministry of Commerce Trade Remedy Public Information Reading Room (Tel: 0086-10-65197878) to search, read, copy, and duplicate the public information of this case.

V. Comments on the Initiation of the Investigation
Interested parties who wish to comment on matters relating to the initiation of the investigation shall submit their written comments to the Bureau of Trade Remedy Investigations of the Ministry of Commerce within 20 days from the date of publication of this announcement.

VI. Submission and Handling of Information
During the investigation, interested parties shall submit electronic versions of comments, questionnaire responses, and other materials through the “Trade Remedy Investigation Information Platform”, and simultaneously submit written versions as required by the Ministry of Commerce. The content of the electronic and written versions shall be identical, and the format shall be consistent.

If an interested party believes that the disclosure of the information it provides would cause serious adverse effects, it may apply to the Ministry of Commerce for confidential treatment and state the reasons. If the Ministry of Commerce approves the request, the interested party applying for confidentiality shall also provide a non-confidential summary of the confidential information. The non-confidential summary shall contain sufficient meaningful information to enable other interested parties to have a reasonable understanding of the confidential information. If a non-confidential summary cannot be provided, the reasons shall be stated. If the information submitted by an interested party does not indicate that it requires confidential treatment, the Ministry of Commerce will treat it as public information.

VII. Contact Information of the Ministry of Commerce
Address: No. 2 East Chang’an Avenue, Beijing
Postal Code: 100731
Bureau of Trade Remedy Investigations, Ministry of Commerce
Tel: 0086-10-65198155 65198070
Fax: 0086-10-65198172
Relevant Website: Sub-website of the Bureau of Trade Remedy Investigations on the Ministry of Commerce website

Ministry of Commerce of the People’s Republic of China
March 27, 2026

The second investigation targets “U.S. practices and measures that hinder trade in green products.” According to the MOFCOM announcement, the United States has implemented a number of practices and measures in trade-related fields that hinder trade in green products, including but not limited to: restricting the export of green products to the United States, slowing down the deployment of new energy projects, and restricting technology cooperation related to green products. These practices and measures may seriously harm the trade interests of Chinese enterprises, and some of them are suspected of violating WTO rules and other economic and trade treaties or agreements jointly concluded or acceded to by China and the United States.

Ministry of Commerce Announcement No. 18 of 2026
Initiation of Trade Barrier Investigation on U.S. Practices and Measures Hindering Trade in Green Products

Issuing Authority: Bureau of Trade Remedy Investigations
Announcement No.: Ministry of Commerce Announcement No. 18 of 2026
Date of Issuance: March 27, 2026

(The remaining sections of the announcement—investigation procedures, period, access to information, comments, submission of information, and contact details—are almost identical to Announcement No. 17, with only the investigated measures changed to: “the practices and measures implemented by the United States in trade-related fields that hinder trade in green products.”)

The above investigations were launched one day after MOFCOM Minister Wang Wentao met with U.S. Trade Representative Jamieson Greer in Cameroon.

According to the MOFCOM readout, during the meeting Wang Wentao “expressed grave concern over the U.S. side’s initiation of Section 301 investigations against multiple economies, including China, on the grounds of so-called ‘overcapacity’ and ‘failure to prohibit the importation of goods produced with forced labor.’”

MOFCOM Readout

Minister Wang Wentao Meets with U.S. Trade Representative Jamieson Greer

On March 26, during the 14th Ministerial Conference of the World Trade Organization held in Yaoundé, Cameroon, Minister of Commerce Wang Wentao met with U.S. Trade Representative Jamieson Greer. The two sides had in-depth exchanges of views on China-U.S. economic and trade relations, as well as multilateral and regional economic and trade cooperation. Ambassador Li Yongsheng of the Chinese Permanent Mission to the WTO attended the meeting.

Wang Wentao stated that President Xi Jinping has pointed out that economic and trade relations should continue to serve as the ballast and propeller of China-U.S. relations, rather than a stumbling block or point of conflict. Both sides should jointly implement the important consensus reached at the Busan meeting and previous calls between the two heads of state, properly handle the relationship between competition and cooperation as well as between the past and the future, strengthen mutually beneficial cooperation, avoid vicious competition, maintain close communication, jointly “look forward,” and promote the healthy, stable, and sustainable development of bilateral economic and trade relations.

Wang Wentao emphasized that China is willing to strengthen multilateral and regional economic and trade cooperation with the U.S. side, jointly advance WTO reform, promote practical outcomes at the 14th Ministerial Conference of the WTO, and support positive results from APEC and G20 meetings. Wang Wentao also expressed grave concern over the U.S. side’s initiation of Section 301 investigations against multiple economies, including China, on the grounds of so-called “overcapacity” and “failure to prohibit the importation of goods produced with forced labor.”

Greer stated that over the past year, the U.S. and China have conducted constructive economic and trade consultations and maintained close communication and cooperation. The U.S. side is willing to work with China, guided by the consensus between the two heads of state, to strengthen dialogue, promote sustained and stable China-U.S. economic and trade relations, and achieve mutually beneficial and win-win outcomes in the economic and trade field. The U.S. side is also willing to strengthen communication and exchanges with China under the WTO and regional frameworks.

Following the U.S. initiation of Section 301 investigations against China, the Chinese Ministry of Foreign Affairs and Ministry of Commerce quickly issued strongly worded responses. They emphasized that the U.S. claim of Chinese “overcapacity” is a false proposition and essentially a “political manipulation” by the U.S. side. They urged the U.S. to correct its erroneous practices and return to the correct track of resolving issues through dialogue and consultation. China also stressed that it will closely follow developments and “reserves the right to take all necessary measures” to resolutely safeguard its legitimate rights and interests.

Statement by the Spokesperson of the Ministry of Foreign Affairs

Reuters: The Trump administration announced on Wednesday that it will launch two new trade investigations into “excess industrial capacity” of 16 major trading partners, including China. According to officials from the Office of the U.S. Trade Representative, this could lead to new tariffs. What is China’s comment on this investigation and on the USTR listing China’s automotive industry as a so-called “overcapacity” sector?

Guo Jiakun: China’s position on handling China-U.S. economic and trade issues is consistent and clear. We oppose unilateral tariff measures in all forms. Tariff wars and trade wars are not in the interest of either side. Both sides should resolve relevant issues through consultation on the basis of equality, respect, and reciprocity. The so-called “overcapacity” is a false proposition, and China opposes using it as a pretext for political manipulation.

Statement by the MOFCOM Spokesperson

Reporter: On March 11 Eastern Time, the Office of the U.S. Trade Representative announced the initiation of Section 301 investigations against 16 economies, including China, on the grounds of “overcapacity.” What is China’s comment?

Answer: China has taken note that the U.S. side has initiated Section 301 investigations against 16 economies, including China, on the grounds of “overcapacity.” Section 301 investigations are typical unilateralist actions that seriously undermine the international economic and trade order. A WTO panel has already ruled that tariff measures taken under Section 301 investigations violate WTO rules.

Regarding the U.S. so-called “overcapacity theory,” China has repeatedly clarified its position. The world economy has long become an indivisible whole; both production and consumption are global, requiring supply-demand matching and adjustment from a global perspective. If production in every country could only meet domestic market demand, there would be no cross-border trade. The U.S. side cannot narrowly define production capacity exceeding domestic demand as “overcapacity” and label it as such. The U.S. has even less right to unilaterally determine through Section 301 investigations whether trading partners have “overcapacity” and then adopt unilateral restrictive measures. China has noted that the U.S. side has also initiated Section 301 investigations against 60 economies, including China, on the grounds of “failure to prohibit the importation of goods produced with forced labor.” China is conducting analysis and assessment.

China urges the U.S. to correct its erroneous practices and return to the correct track of resolving issues through dialogue and consultation. China will closely follow developments, reserve the right to take all necessary measures, and resolutely safeguard its legitimate rights and interests.

In the recent China-U.S. talks in Paris, Vice Premier He Lifeng also expressed great concern over the U.S. Section 301 investigations.

Recently, the U.S. Supreme Court ruled that tariffs imposed under the International Emergency Economic Powers Act (IEEPA) were unlawful. Subsequently, the United States imposed a 10 percent import surcharge on all trading partners under Section 122 of the Trade Act of 1974 and has also introduced a number of negative measures targeting China, including Section 301 investigations, corporate sanctions, and market access restrictions.

China’s opposition to the United States’ imposition of unilateral tariffs has been consistent. China urges the United States to completely remove unilateral tariffs and other restrictive measures and will take necessary actions to firmly safeguard its legitimate rights and interests.

In accordance with the relevant provisions of China’s Foreign Trade Law and the Ministry of Commerce’s Rules on Foreign Trade Barrier Investigations, after initiating an investigation into trade barriers of relevant countries and regions, the Ministry of Commerce may use questionnaires, hearings, on-site investigations, and other methods to understand the situation from interested parties and conduct the investigation. It will form different conclusions based on the actual investigation results, which will lead to corresponding legal and practical consequences.

During the investigation, if the party under investigation voluntarily cancels or modifies the relevant restrictive measures, provides reasonable compensation, or fulfills its obligations under relevant international agreements, the investigation may be suspended or even directly terminated in accordance with the law, and subsequent determination procedures will no longer be pursued. If the investigation ultimately confirms that the relevant measures of the country or region do constitute unreasonable restrictions and harm to the entry of Chinese goods, services, or investment into its market, they will be formally determined to constitute trade barriers.

Once a trade barrier is determined to exist, MOFCOM, as appropriate, may adopt one or more of the following measures in response in accordance with Article 33 of the Rules:

First, engage in bilateral official consultations with the government of the country or region under investigation, urging it to remove or modify the relevant barrier measures or provide trade compensation, so as to resolve the dispute in a friendly manner;

Second, initiate dispute settlement procedures through multilateral international platforms such as the World Trade Organization (WTO) dispute settlement mechanism, file complaints, and promote the issuance of rulings, requiring the relevant country or region to correct its non-compliant measures;

Third, take other appropriate measures, including but not limited to imposing reciprocal trade restrictions on the country or region under investigation, such as additional tariffs or restrictions on market access; issuing risk alerts to guide Chinese enterprises in avoiding relevant market risks; supporting affected enterprises in filing applications for trade remedy measures such as anti-dumping and countervailing duties; and coordinating with other countries or regions to jointly exert pressure in order to eliminate the relevant trade barriers.

If, after comprehensive verification, it is determined that the relevant measures do not constitute trade barriers or that the existing evidence is insufficient to establish the existence of trade barriers, the investigation procedure will be normally terminated, the relevant conclusions will be publicly announced, and in principle, no repeated investigation will be conducted on the same grounds. Relevant market entities will then need to continue conducting foreign economic and trade activities in accordance with existing rules.

According to reports citing sources familiar with the discussions, the talks were described as “remarkably stable” and “candid and constructive.” They are also expected to help prepare potential “deliverables” for the meeting between the two countries’ leaders scheduled for the end of March. In that sense, the talks were clearly more than a routine working-level exchange; they were aimed at laying the groundwork for the potential leaders’ meeting and identifying possible outcomes in advance.

Based on public statements from both sides, the agenda of the talks included the following items:

1. Tariffs: the United States’ new Section 122 tariffs and two ongoing Section 301 investigations.

2. Trade: agricultural purchases (including soybeans, poultry, beef, and certain non-soybean crops), purchases of Boeing aircraft, and energy exports (coal, oil, and natural gas), as well as the possible establishment of a Board of Trade.

3. Investment: the proposal to establish a Board of Investment.

4. Non-tariff measures: China’s export controls on rare earths, particularly yttrium, a heavy rare earth element critical for jet engine turbines. Greer noted that U.S. companies occasionally report difficulties obtaining rare earth supplies, so the two sides discussed the implementation of the rare-earth provisions under the Busan agreement. This may suggest that some U.S. companies are still encountering supply bottlenecks and have raised the issue with the U.S. government.

China released a readout via Xinhua News Agency.

Non official translation:

Xinhua News Agency, Paris, March 16 — From March 15 to 16 local time, China’s lead representative for China–U.S. economic and trade affairs, Vice Premier He Lifeng, held economic and trade consultations in Paris, France, with the U.S. lead representatives, U.S. Treasury Secretary Scott Bessent and U.S. Trade Representative Jamieson Greer.

Guided by the important consensus reached by the two heads of state, the two sides conducted candid, in-depth, and constructive exchanges on economic and trade issues of mutual concern, including tariff arrangements, the promotion of bilateral trade and investment, and the maintenance of previously reached consultation outcomes. The discussions produced some new consensus, and the two sides agreed to continue maintaining consultations.

He Lifeng stated that under the strategic guidance of the important consensus between the two presidents, the two sides had reached a series of outcomes in the economic and trade field through five rounds of consultations last year, injecting greater certainty and stability into China–U.S. economic relations and the global economy.

Recently, the U.S. Supreme Court ruled that tariffs imposed under the International Emergency Economic Powers Act (IEEPA) were unlawful. Subsequently, the United States imposed a 10 percent import surcharge on all trading partners under Section 122 of the Trade Act of 1974, and has also introduced a number of negative measures targeting China, including Section 301 investigations, corporate sanctions, and market access restrictions.

China’s opposition to the United States’ imposition of unilateral tariffs has been consistent. China urges the United States to completely remove unilateral tariffs and other restrictive measures, and will take necessary actions to firmly safeguard its legitimate rights and interests.

He also expressed hope that the United States will work with China in the same direction, jointly implement the important consensus reached by the two heads of state, expand the list of cooperation while narrowing the list of problems, and promote the healthy, stable, and sustainable development of China–U.S. economic and trade relations.

The U.S. side stated that stable China–U.S. economic and trade relations are very important for both countries and for the world, and contribute to global economic growth, supply chain security, and financial stability. Both sides should reduce frictions, avoid escalation of disputes, and resolve differences through consultation.

The two sides agreed to study the establishment of cooperation mechanisms to promote bilateral trade and investment, continue to make good use of the China–U.S. economic and trade consultation mechanism, strengthen dialogue and communication, properly manage differences, expand practical cooperation, and promote the sustained and stable development of bilateral economic and trade relations.

US readout coming out later:

In advance of @POTUS’ trip to Beijing, @USTradeRep Jamieson Greer and I met with Vice Premier He Lifeng of China to discuss areas of mutual interest.

Talks were candid and constructive, and we are on a good path towards the next meeting between President Trump and General Secretary Xi.

In our discussions, we addressed a number of priorities to promote stability in our relationship, including how to manage bilateral trade and improve market access for the United States.

Both sides subsequently briefed the media. U.S. Treasury Secretary Scott Bessent also gave an interview to CNBC.

Key points from Bessent’s CNBC interview

1. Whether the presidential visit to China will proceed as scheduled remains to be seen. However, Bessent clarified that media reports suggesting a possible delay because the president asked China to patrol or enforce security in the Strait of Hormuz are completely inaccurate. If the visit were postponed, it would likely be due to logistical considerations, as the president may wish to remain in the United States to coordinate wartime operations, making overseas travel less appropriate at such a time.

2. His meeting with Vice Premier He Lifeng went very smoothly. In the coming days, China and the United States are expected to issue a statement reaffirming the stability of bilateral relations.

3. Semiconductor export controls were not discussed. The meeting focused primarily on economic matters, including potential Chinese purchases of Boeing aircraft and agricultural products, while the U.S. side briefed China on the new Section 122 tariffs and the ongoing Section 301 investigations.

CNBC

Thank you very much. We are here in Paris at the OECD headquarters with the Treasury Secretary of the United States. God bless you, Mr. Treasury Secretary. Thank you very much for spending some time with CNBC.

Well, you literally just moments ago wrapped up meetings with the Chinese trade delegation. Tell us how the meetings went and whether or not we should expect President Trump to meet with China in April.

Bessent

The meetings were very good. We’ve got a stable relationship. This is our sixth meeting in our economic consultations. The Vice Premier, He Lifeng, and I have developed great respect for each other. But really, the meetings are generated by the great respect that the two leaders present — President Trump and Party Chairman Xi — have for each other.

You know, we will see whether the visit takes place as scheduled. But what I do want to address is that there’s a false narrative out there suggesting that if the meetings are delayed, it would be because the president demanded that China police the Strait of Hormuz.

CNBC

There have been reports out there to that effect.

Bessent

Yes — that one is completely false. If the meeting were for some reason rescheduled, it would be because of logistics. The president wants to remain at the NSC to coordinate the war effort, and traveling abroad at a time like this may not be optimal.

CNBC

Because that narrative is out there. So you’re not saying that the meeting will be rescheduled, but you’re saying — and correct me if I’m wrong — if it is rescheduled, it would be because of timing and travel, not because of a dispute over the Strait.

Bessent

Exactly. It would be a decision the president makes as Commander-in-Chief to stay in the White House or remain in the United States while this war is being prosecuted.

CNBC

Because you’re a markets expert, obviously a long-time market participant. The markets may react to any word of a delay or pause in the meeting. And you’re saying if this meeting is paused or delayed, the market should not interpret it as a conflict between the two nations.

Bessent

Absolutely not. We had a very good two days here. We’ll be issuing a statement in the next few days reaffirming the stability of the relationship between the first- and second-largest economies in the world.

CNBC

I want to get to the war, I want to get to oil, I want to get to commodities. But first I want to finish up on this meeting with China. It sounds positive.

Were the discussions about things like export controls on advanced American semiconductors? Will the Chinese agree to buy more Boeing jets or other U.S. goods? Can you go into a little more detail about the products and services that were discussed here?

Bessent

Sure. We really don’t discuss our export restrictions in these meetings — it’s more of an economic dialogue. What we do discuss, exactly as you said, are purchase commitments from the Chinese.

We also explain to them our new tariff regime and how that would work — the Section 122 tariffs that were implemented right after the Supreme Court ruled against the president’s use of IEEPA.

And we explain the Section 301 studies that Ambassador Greer and USTR will be conducting, which should be completed by July.

CNBC

Is there an opportunity — assuming the market reaction matters — the stock market is down about 5% from its highs. It hasn’t collapsed, but it is down. Markets are nervous. If you look at the VIX, the volatility index, it’s elevated.

Would you ever consider using tariffs — the 15% tariffs — as a bargaining chip with China vis-à-vis the markets? What are we bargaining for？

Bessent

Because the president has made it very clear. Over the past year, President Trump — through USTR, through myself, and through Commerce — we’ve negotiated trade deals with most of our major trading partners and reordered global trade.

Our goods trade deficit with the world is down. Our goods trade deficit with China is down.

We had substantial tariff income last year. The court said we’ll have to refund that. But I can tell you that because of the change from IEEPA to Section 122, and we’ll see what the Section 301 studies bring — assuming a successful completion of those — it’s very likely that Treasury will see very little change in tariff revenue.

Again, the ultimate purpose of tariffs is to reshore production. The revenues are a very nice incidental benefit along the way. In a sense, it’s payback for the offshoring that occurred because of unfair trade practices.

But at the end of the day, we want to reshore manufacturing in the United States, especially strategic manufacturing.

And to go back to our consultations with the Chinese: we reiterated to them that we do not want to decouple, but we do need to take strategic industries back.

So we are making a strategic adjustment, but not pursuing a generalized decoupling in trade.

Key points from the media briefing by Bessent and U.S. Trade Representative Jamieson Greer

1. Regarding whether President Trump’s visit to China might be postponed, the U.S. side stated that any delay would not be due to China failing to meet U.S. requests. Rather, it would more likely relate to broader issues such as compliance matters, the status of trade agreements, and overall economic and trade arrangements between the two countries.

2. The two sides are currently sorting through a number of possible items that could be submitted to the two leaders. A key focus is to further clarify and institutionalize the structure of bilateral trade, including which products the United States should import from China and which products it should export to China, with the goal of making bilateral trade more balanced and more focused on mutually beneficial sectors.

Please note that the following content is based on a transcription of unclear audio and may differ from the speaker’s original remarks.

Reporter:

This meeting was intended to help finalize some potential deliverables for the two presidents to consider when they meet in Beijing next month. Could you give us an idea of what those might be going forward? And given what the president has said about his schedule, is that meeting still expected to take place at that time?

Bessent:

If the president’s visit is postponed, it would have nothing to do with the Chinese making a commitment via to the strengths of Hormuz. Uh it would obviously be in their interest to do so, but a postponement would not be as a result of any ask from the president not being met. The postponement if it happens would be because the commander-in-chief of the United States military believes that he should stay in the United States while this war is being prosecuted.

Greer:

First what we’ve concluded today really is the general terms of a work plan between now and the meeting between the presidents with the idea that there will be potential deliverables at that meeting.

We of course discussed compliance with the Busan agreement. This means rare earths etc. Uh you know from time to time we will get information from US companies or stakeholders about the status of receiving rare earths. And so we covered issues like that related to the agreement. We also talked about expanding trade exports to China in terms of agricultural goods, energy goods etc.

Reporter:

Have you been made aware of any China’s retaliation related Section 301 investigation?

Bessent:

Well, there there were very detailed discussions in terms of the the new tariff authorities. We will be implementing uh it’s the 122. We discussed the USTR and Ambassador Greer’s investigations. But in terms of of retaliation, as I said, there’s great stability in the relationship and the the purpose of these meetings is to prevent any retaliation as we saw this time last year.

Greer:

And I would say we started these talks really by giving them a a preview of what we’re doing on US trade policy as we adjust to the Supreme Court.

Remember, the president’s trade policy hasn’t changed. Our tools may change and we’re conducting these investigations.

We don’t want to prejudge them. And we had a good conversation with our counterparts about that process.

Reporter:

Just briefly — is there a possibility that the president might not go to Beijing? Are you suggesting that as long as the war continues, the commander-in-chief might reconsider the trip?

Bessent:

That’s not what I said. I want to make it clear that our position remains consistent.

Key points from the media briefing by China’s chief trade negotiator Li Chenggang

1. The discussions covered several topics, including tariff levels under the new circumstances, the possible extension of bilateral tariff and non-tariff measures, and ways to promote bilateral trade and investment cooperation.

2. Regarding tariff levels under the current circumstances, the U.S. side briefed China on its recent tariff adjustments and its considerations for future measures. The Chinese side expressed concern about the uncertainty created by these developments. Both sides agreed to continue working toward maintaining the stability of bilateral tariff levels, and they also held in-depth discussions on each other’s economic and trade concerns.

3. The two sides reviewed the implementation of the outcomes of the five rounds of economic and trade consultations held last year, giving overall positive recognition to the progress made. At the same time, the Chinese side made solemn representations regarding the two ongoing Section 301 investigations, expressing serious concern. China will closely monitor the progress of these investigations and take appropriate measures to safeguard its legitimate rights and interests.

4. The two sides have reached preliminary consensus on some issues and will continue to maintain the consultation process going forward.

Li Chenggang’s remarks:

Friends from the press, good afternoon. Thank you for your attention.

First of all, I would like to thank the French government for its hospitality, and also express our appreciation for the strong support from the OECD.

This time we are here in Paris to engage in a new round of consultations under the China–U.S. economic and trade dialogue.

Over the past one and a half days, the two teams have carried out in-depth, candid and constructive consultations.

The topics we discussed include tariff levels under the new circumstances, the bilateral trade situation, the possible extension of bilateral tariff and non-tariff measures, as well as the promotion of bilateral trade and investment. Both sides also exchanged views on each other’s economic and trade concerns.

Regarding tariff levels under the current circumstances, U.S. colleagues provided us with information about their recent tariff measures and their relevant considerations going forward. The Chinese side shared its concerns about the uncertainties created by these measures.

Both sides agreed to remain committed to maintaining the stability of bilateral tariff levels.

On the issue of promoting bilateral trade and investment, the two sides discussed the idea of establishing a working group to study possible cooperation mechanisms aimed at facilitating bilateral trade and investment.

During the consultations, both sides also had in-depth discussions on their respective economic and trade concerns.

Together, the two sides reviewed the implementation of the outcomes of the five rounds of economic and trade consultations held last year and gave overall positive recognition to the progress made.

At the same time, we have noticed that the United States has introduced a number of restrictive measures concerning China’s trade and investment.

In particular, the recent launch of two Section 301 investigations includes China among the targets.

The Chinese side made solemn representations to the U.S. side during the consultations and expressed our serious concerns.

China’s position on the Section 301 investigations has been consistent. We oppose such unilateral investigations.

We are also concerned about the potential interference and damage that the results of these investigations may cause to the hard-won stability of China-U.S. economic and trade relations.

China will continue to closely follow the development of these investigations and will take timely measures to safeguard China’s legitimate rights and interests.

During this round of consultations, both sides further recognized that a stable China-U.S. economic and trade relationship is beneficial to both countries and to the global economy.

We hope that the U.S. side will act in good faith, honor its commitments, and work with China in the same direction to promote the long-term, steady development of bilateral economic and trade relations.

Through this round of consultations, the two sides have reached preliminary consensus on certain issues. Moving forward, both sides will continue to maintain the consultation process.

Thank you for your continued attention.

Based on information that has emerged in media reporting, the discussions between the two sides appear to have focused on several main areas.

1. Agricultural Trade

The U.S. side has long sought to expand its agricultural exports to China. During this round of talks, the Chinese side showed a degree of openness to increasing purchases of certain U.S. agricultural products. The items discussed reportedly included poultry, beef, and several non-soybean crops.

At the same time, China reiterated its commitment to maintain purchases of roughly 25 million metric tons of U.S. soybeans annually over the coming years.

2. Establishing a New Institutional Framework for U.S.–China Economic Dialogue

The two sides appear to have discussed a more institutionalized framework to manage trade and investment relations over the longer term. The basic concept involves creating two standing mechanisms: one focused on trade and the other on investment.

The proposed “Board of Trade” appears to be the more developed concept. Its basic idea is to identify products and sectors where the two countries could expand trade without compromising national security or critical supply chains. In essence, it would seek to expand cooperation in areas considered non-sensitive, while avoiding sectors with strategic implications.

The proposed “Board of Investment” would function somewhat differently. Rather than setting broad investment policy, it would likely serve as a platform for addressing specific investment disputes or practical issues encountered by companies operating in each other’s markets. In other words, its purpose would be to deal with “discrete investment issues”, rather than to fundamentally alter the existing investment review regimes on either side.

3. Supply of Critical Minerals (Especially Rare Earth Elements)

The U.S. side also raised the issue of American companies’ access to critical minerals produced in China, including yttrium, a rare earth element used in applications such as jet engine turbines.

The U.S. aerospace industry has been particularly concerned about supply constraints for this material. According to some accounts, the two sides may have identified potential ways to ease tensions surrounding critical mineral supply, although the details have not been publicly disclosed.

Further technical-level consultations are expected to continue. Discussions are likely to focus on the two proposed institutional mechanisms mentioned above, as well as specific U.S. requests for expanded Chinese purchases of Boeing aircraft and American energy products such as coal, oil, and natural gas.

4. The Concept of “Managed Trade”

In recent months, U.S. officials have increasingly discussed the idea of “managed trade.” This concept also involves distinguishing between “sensitive” and “non-sensitive” goods.

In December of last year, U.S. Trade Representative Jamieson Greer offered a relatively comprehensive explanation of this approach. In his view, the United States should actively promote exports in sectors where it holds a competitive advantage, such as agricultural products, aircraft, and medical equipment.

Consumer goods and low-technology products could continue to circulate under normal market-driven trade frameworks. However, high-technology sectors inherently involve national security considerations and therefore cannot be left entirely to market forces. Instead, they require government-level scrutiny to ensure that U.S. national security and technological leadership are not compromised.

Before departing for the Paris talks, Greer reiterated in an interview with CNBC that the United States hopes to maintain a stable relationship with China and achieve more balanced trade, but that bilateral trade should focus primarily on non-sensitive goods.

5. Investment Issues and Domestic Political Constraints in the United States

Regarding the proposed investment mechanism, the United States has already established a relatively robust framework for reviewing Chinese investment. This system is built around legislation such as the Foreign Investment Risk Review Modernization Act (FIRRMA) and related policy instruments.

Given the political pressure from Congress and national security hawks, significant changes to this framework appear unlikely in the near term.

At the same time, reports suggest that China has explored broader cooperation ideas, including the possibility of sovereign-capital-backed funds investing in certain U.S. industries, such as batteries or electric vehicles.

Although Trump has previously said he would welcome Chinese investment in electric vehicles, such proposals face substantial domestic opposition in the United States. The House Select Committee on the Chinese Communist Party has written to the Treasury Department urging it to block Chinese EV entry into the U.S. market. Several major U.S. auto industry associations have also called on the government to restrict Chinese companies from entering the American automotive sector, including through local manufacturing.

Under such circumstances, focusing cooperation on specific investment issues may represent a more realistic path forward.

6. Prospects for Reviving Bilateral Investment

According to some reports, working-level officials from both countries have begun exploring ways to revive mutually beneficial investment. Areas of potential interest reportedly include structured joint ventures, licensing arrangements, and so-called “light intellectual property” cooperation models.

The Chinese side has emphasized the need for stronger investment protection, including greater clarity regarding tariffs affecting supply-chain components and intermediate inputs. The U.S. side, meanwhile, continues to stress managed trade principles and reciprocity.

Chinese officials have also expressed concern about the increasingly stringent scrutiny applied to Chinese investment in the United States, noting that such reviews have already led to declining investment flows and, in some cases, withdrawals.

Meanwhile, the Committee on Foreign Investment in the United States (CFIUS) is reportedly reviewing whether Tencent’s investments in certain gaming companies in the United States and Finland pose national security risks. According to reports, some officials believe mitigation measures could address the concerns, while others argue that the risks are too significant, particularly because gaming companies collect large volumes of user data, including financial information, personal profiles, and chat records.

7. Possible Uncertainty Around Trump’s Planned Visit to Beijing

In a media interview on Sunday, Donald Trump indicated that he has been urging China to play a greater role in ensuring the security of the maritime route through the Strait of Hormuz. If discussions on that issue do not make progress, he said he would not rule out adjusting or postponing the planned visit.

According to his reasoning, compared with the United States, Europe and China rely more heavily on oil shipments from the Gulf region, and therefore countries that benefit from that shipping route should contribute to maintaining its security.

China’s MOFA responds to this, as well as to how to deal with Rubio’s China sanctions if he is to visit Beijing along with Trump. My general take from it is: 1)Beijing will not send ships to the Strait of Hormuz; 2) China is hoping Rubio’s visit and the sanctions are not an issue anymore.

Anadolu Agency reporter:

U.S. President Donald Trump said he is in talks with seven countries about forming an escort coalition in the Strait of Hormuz, which could be announced as early as this week. Trump also wrote on social media that he hopes countries affected by a potential closure of the Strait of Hormuz — including China, France, Japan, South Korea, and the United Kingdom — will dispatch naval vessels to the area to ensure the strait remains secure and free from threats. What is China’s comment on this? Has the U.S. side asked China to participate in the proposed escort coalition?

Lin Jian:

The situation in the Strait of Hormuz and surrounding waters has recently become tense, disrupting international trade routes for goods and energy and undermining regional and global peace and stability. China once again calls on all parties to immediately cease military actions, avoid further escalation, and prevent instability in the region from causing greater negative impacts on global economic development.

EFE reporter:

Some media outlets have reported that U.S. Secretary of State Marco Rubio may travel to China later this month together with President Trump. Since Rubio has been under Chinese sanctions since 2020, will those sanctions affect his possible visit to China?

Lin Jian:

China’s sanctions target Mr. Rubio’s China-related statements and actions during his tenure as a U.S. Senator.

Overall impressions

First, Trump’s visit to China could potentially be postponed.

However, as Treasury Secretary Scott Bessent repeatedly clarified, a delay would not be related to pressuring China to deploy naval vessels to escort shipping in the Strait of Hormuz. Rather, the reason would be that the situation surrounding the Iran war has recently become quite tense, and Trump may need to remain in the United States to coordinate wartime operations, making it difficult for him to travel abroad.

Trump’s remarks to the Financial Times last Sunday—that if China does not play a greater role in ensuring the security of the Strait of Hormuz he would not rule out adjusting or postponing the planned visit—were most likely a form of strategic pressure. It is also possible that those comments were made without prior coordination with cabinet officials such as Bessent. To some extent, Bessent has been trying to walk those remarks back.

At the very least, however, Trump’s statement indicates that—whatever the reason—the White House may indeed be considering whether the visit should be delayed.

Second, the U.S. side appears to prefer framing this round of talks as primarily an “economic dialogue,” while avoiding or sidestepping broader geopolitical or security issues.

One example came when a CNBC reporter asked Bessent whether the talks included discussions on U.S. semiconductor export controls. Bessent answered very clearly: U.S. export controls were not discussed at this meeting, emphasizing that the discussions were meant to function more as an economic dialogue.

Another example comes from China’s Vice Minister of Commerce and Chief Trade Negotiator Li Chenggang, who stated in his media briefing that the topics discussed included:

• tariff levels under the new circumstances

• the possible extension of bilateral tariff and non-tariff measures

• promoting bilateral trade and investment cooperation

Here, the “non-tariff measures” referred to China’s export controls on rare earths. Notably, semiconductor export controls were not mentioned, suggesting the issue was not part of the formal agenda.

At the same time, Reuters reported on the same day—citing sources familiar with the matter—that Hua Hong Semiconductor is now capable of producing 7-nanometer AI chips, and its subsidiary HLMC (Huali Microelectronics) is preparing to launch a 7-nm process line at its Shanghai facility. This would make Hua Hong the second Chinese foundry after SMIC capable of producing 7-nm AI chips.

The report also indicated that Huawei has been working with Hua Hong on 7-nm technology, supported by domestic suppliers including SiCarrier, and that Biren Technology, a GPU design company on the U.S. Entity List, has already begun taping out chips using Hua Hong’s 7-nm production line.

If this trend continues, China’s incentive to raise the issue of semiconductor export controls in negotiations with the United States may gradually diminish over time.

Third, the Chinese side paid particular attention to two newly launched U.S. Section 301 investigations.

The first investigation focuses on structural overcapacity in foreign manufacturing and related industrial policies, while the second targets trade practices related to forced labor. In essence, these investigations represent part of the United States’ effort to rebuild its trade pressure toolkit after the “reciprocal tariffs” were struck down.

However, these investigations also introduce new uncertainty into the tariff arrangements previously reached between China and the United States, since no one yet knows what level of tariffs could ultimately result from these cases.

In response to this uncertainty, the Chinese side made solemn representations during the consultations and expressed serious concerns.

As Li Chenggang stated:

China’s position regarding these investigations has been consistent. We oppose such unilateral investigations. We are concerned that the possible outcomes of these investigations may interfere with and undermine the hard-won stability of China–U.S. economic and trade relations. We will closely monitor the progress of these investigations and take appropriate measures to safeguard China’s legitimate rights and interests.

The U.S. side also spent considerable time explaining the Section 122 tariffs and the two Section 301 investigations.

According to U.S. Trade Representative Jamieson Greer, the U.S. delegation began the talks by briefing the Chinese side on the fact that the U.S. Supreme Court had struck down the reciprocal tariffs, and on the subsequent launch of the two Section 301 investigations. He emphasized that U.S. trade policy itself has not changed; only the policy tools being used have changed.

Greer also noted that the investigations are still ongoing and that the United States does not want to prejudge the outcomes. The two sides held constructive discussions about the investigative process.

When asked by reporters whether China had threatened retaliation, Bessent responded that China–U.S. relations remain broadly stable, and that one of the purposes of these meetings is precisely to prevent the kind of retaliatory escalation seen at this time last year.

If Trump proceeds with the visit to China at the end of the month as originally planned, this Paris meeting would likely be the final round of talks before the summit. However, compared with past summits, the preparatory pace has been far less intensive, and many technical-level negotiations have not yet reached the stage where details can be finalized.

The Iran war and broader Middle East tensions are also introducing uncertainty around the summit, and discussions related to Iran could potentially consume valuable time during the leaders’ meeting.

That said, it may be better not to place excessively high expectations on a single summit at the end of the month. Looking at the broader timeline of China–U.S. engagement this year, there are still two major opportunities for high-level interaction:

• APEC in Shenzhen in November

• G20 in Miami in December

The potential Beijing meeting may therefore serve more as a stepping stone for future outcomes, rather than a venue where major issues are resolved all at once.

The Paris talks also suggest that both sides are likely to continue advancing two new channels of bilateral dialogue—the proposed Board of Trade and Board of Investment—over the coming months. These mechanisms could be used to explore issues such as:

• investment opportunities for Chinese companies in the United States

• arrangements for bilateral investment cooperation

• operationalizing the U.S. concept of “managed trade” and “non-sensitive goods”

• restoring normal trade in selected product categories

At the upcoming APEC and G20 meetings, the two sides are also expected to discuss a range of important issues concerning the broader U.S.–China relationship.

As for the Section 301 investigations, China is clearly one of the targets. However, because of China’s leverage in rare earth supply chains, the U.S. side is unlikely to push too aggressively. This round of discussions therefore appeared to focus largely on explaining, clarifying, and reassuring the Chinese side. At the same time, China made its position clear from the outset, drawing red lines and issuing warnings about the potential consequences.

According to Treasury Secretary Scott Bessent, the two Section 301 investigations will not be completed until July, which means both sides still have plenty of time to negotiate and manage the issue.

As for the Section 122 tariffs, their impact on China is likely to be minimal. First, the measure only has a 150-day window, and it is already facing legal challenges from Democratic state governments and U.S. businesses in domestic courts.

Moreover, under the Section 122 tariff framework, the average U.S. tariff level on Chinese goods has effectively been reduced by about 10 percentage points, which provides China with a meaningful buffer.

How should we assess the outcome of these talks?

I think expectations were actually quite high going in. But the working-level teams on both sides had probably already lowered expectations beforehand, so the positive tone we’re hearing now mostly reflects that the outcome landed pretty much within what both sides expected.

From the U.S. side, a key takeaway is that they’ve gotten some level of reassurance from China on a few issues — things like agricultural and energy purchases, Boeing orders, and rare earths. That doesn’t mean immediate implementation, but at least the direction is aligned.

On the Chinese side, this round helped them get a clearer sense of where the U.S. stands on things like the Section 301 investigations. They now have a better read on the boundaries. On trade and investment, neither side really made concrete concessions, but they did agree to set up mechanisms and outline a path forward — which still matters from a policy perspective.

Overall, both sides got something, but it’s all fairly preliminary. No one walked away with everything they wanted, but both have enough to show for it.

More importantly, this round basically kicks off this year’s U.S.–China engagement, and the start has been relatively stable.

If Trump does go ahead with the planned visit to Beijing later this month, the Paris talks will likely be the last round of consultations before that. But compared to past summits, the preparation this time doesn’t look nearly as intensive. A lot of the technical discussions are still at the framework stage, not yet at the point where details can be finalized.

At the same time, the Iran war and the broader Middle East situation add another layer of uncertainty. If things escalate, Iran could take up a significant portion of the leaders’ agenda, squeezing out time for bilateral issues.

So rather than expecting too much from the Beijing meeting, it probably makes more sense to look at U.S.–China engagement over a longer timeline. There are at least two major touchpoints later this year — APEC in Shenzhen in November and the G20 in Miami in December. In that sense, the Beijing meeting is more likely to serve as a setup for future outcomes, rather than delivering major breakthroughs on its own.

The Paris talks also suggest that both sides are converging on the idea of building new mechanisms to support trade and investment cooperation. In practice, that likely means setting up a Board of Trade and a Board of Investment as ongoing channels, where more specific issues can be worked through over the coming months — including things like Chinese investment in the U.S. and related arrangements.

At the same time, both sides may try to operationalize ideas like “managed trade” and “non-sensitive goods,” and explore ways to resume or expand normal trade in selected categories. These issues will likely continue to be discussed in upcoming multilateral settings like APEC and the G20.

When asked which of the three factors—compute, data, or algorithms—is the biggest bottleneck for AI development, and which breakthrough could drive the next wave of progress, he said the answer changes over time.

In the early days, many people believed algorithms were the key, and that advances in algorithms could take AI directly toward AGI. Later, as model sizes and parameter counts expanded dramatically, people began to worry that the amount of usable data on the internet might be insufficient, raising concerns about “running out of data” and hitting a ceiling in pre-training.

As researchers started exploring ways to address the data problem, it gradually became more manageable. At that point, the focus shifted to compute, with concerns that there might not be enough computational power to sustain continued scaling. But with technological progress and companies like NVIDIA ramping up production aggressively, access to compute has also improved.

Now the conversation is shifting again. Some people argue that the algorithms themselves may be the limiting factor—that current methods may be inefficient, struggle with issues such as catastrophic forgetting, and may not scale indefinitely. Many researchers even predict that the Transformer architecture itself could eventually be replaced or fundamentally redesigned.

He argues that AGI does not currently have a single authoritative or clearly defined meaning. Compared with “AI,” the conceptual boundaries of AGI remain ambiguous. Both academia and industry hold different interpretations of its scope and implications. As a result, any discussion about AGI inevitably begins with the question of how it should be defined.

In his view, AGI is likely achievable as long as the definition is reasonable. If AGI is defined in a way that is not overly grand or unrealistic, but instead operational and concrete, then achieving it becomes more a matter of time rather than principle. From the day it was founded, Zhipu has set AGI as its long-term goal and has developed its own internal framework for defining it.

AGI is often defined using behavioral criteria, similar to how the Turing Test defines intelligence through observable behavior. One common view is that when AI reaches the average level of human performance across a wide range of capabilities, it can be considered AGI. Once such a behavioral definition is adopted, it becomes possible to work backward to determine the technical roadmap—gradually improving models so that their performance across different tasks approaches or matches human-level ability.

When asked whether the world’s main general-purpose large model players — now roughly four or five, mostly in China and the United States, with Europe’s Mistral as another example — might eventually narrow down to just one, Zhang Peng said his instinct was that this would not happen.

In his view, technological development needs diversity, especially in the early and middle stages, when there are still many possible paths for innovation and many different research directions. At that stage, companies will naturally differentiate from one another, and each can still find room to survive. On top of that, the market itself is very large and still expanding rapidly, so there is enough space for multiple players and no reason for the field to converge too quickly.

He argued that in the early and middle phases, the ecosystem is likely to remain diverse. Even if one company gains an advantage, that does not mean it will automatically take the whole market. A true winner-takes-all outcome is more likely only once the technology becomes relatively stable and the pace of innovation slows, at which point Matthew effects become stronger. At the current stage, however, he does not think one company dominating everything is likely, nor does he see any clear sign that the industry is heading in that direction.

Transcript of the conversation is below:

The public’s questions about AI still have not gone beyond Turing’s nine objections

Fang Sanwen: What is AI?

Zhang Peng: AI is an abbreviation for Artificial Intelligence. In the most straightforward sense, it means using technological methods — whether computers or other means — to simulate human intelligence and ultimately serve people.

Fang Sanwen: In his 1950 paper Computing Machinery and Intelligence, Turing proposed the concept of intelligence, and later the Dartmouth conference directly defined artificial intelligence. What is the relationship between the AI we talk about today and Turing’s concept of intelligence, or the AI defined in 1956?

Zhang Peng: If we describe it in terms of logic or mathematics, intelligence is the bigger circle, and artificial intelligence is one part of it — specifically, our attempt to simulate human intelligence. So when we talk about AI or artificial intelligence today, we are still basically referring to the concept proposed at the 1956 Dartmouth conference. That said, as time has passed and technology and markets have changed, the meaning of AI has continued to evolve. Today’s AI covers a broader range of things than it did back then, but the core objective has not changed.

Fang Sanwen: Are large models and AI the same thing?

Zhang Peng: Large models are just one of the technical approaches we use to realize artificial intelligence.

Fang Sanwen: From the 1950s to today, what major milestones has AI gone through?

Zhang Peng: It has been more than 70 years, though not yet 80. Over those decades, AI development has not been smooth at all. Most people agree there have been three waves of AI, or strictly speaking, you could say we are now in the fourth. Why has it risen and fallen? I think that is historically inevitable. Nothing develops in a perfectly upward curve forever. Along the way you run into all kinds of challenges and difficulties, and when combined with the broader social and economic context of the time, ups and downs are perfectly normal.

Fang Sanwen: What was the perceptron proposed in 1958?

Zhang Peng: The perceptron is a mathematical method. Put simply, it tries to find a so-called “hyperplane” in the space of data, dividing that space into two parts: one side corresponds to the result we want, and the other to what we do not want. In essence, it is a classification problem. The perceptron is one method for finding that hyperplane. You set up a model and let it locate the hyperplane more accurately through iterative learning from data. Its significance for AI is that it laid an important foundation: it introduced the idea that machines can learn from data to solve specific problems. Just as humans learn through practice and refine their understanding through feedback, the machine iterates toward a solution rather than applying a fixed formula once and for all. That is fully consistent with the logic behind today’s large models and deep learning. You could say it is the origin, the ancestor, of machine learning.

Fang Sanwen: In 1966, MIT developed what is generally regarded as the first true chatbot, Eliza. Today people are especially fascinated with ChatGPT, which is also all about conversation. Why is AI so closely tied to chatting?

Zhang Peng: I think it comes back to the essential goal of artificial intelligence: using machines or technology to achieve something like human intelligence. But how do you determine whether what you have built really has human-like intelligence? You need some way to test it. For humans, the most natural test is conversation: can I tell whether you are a person or not? That starting point shaped a lot of the work that came later.

Fang Sanwen: In 1973 there was the Lighthill Report, which highlighted the limitations of AI and led to a decline in investment. What were those limitations at the time?

Zhang Peng: The Lighthill Report pointed to several things. First, the AI research community at the time was dominated by optimism. Even back then people were already proclaiming that we would achieve general artificial intelligence within 20 years. The vision was admirable and the goals were ambitious, but the practical difficulties were huge. At the time, computers had very weak processing power. Second, data was not well organized. Third, early AI methods were basically based on the symbolic school, which could only solve a very limited class of problems — things like mathematics and physics that could be expressed in a complete symbolic system. The moment you expanded into broader knowledge or common sense, it stopped working. So people began to reassess AI investment more soberly, and that led to the first AI winter.

Fang Sanwen: Is it fair to say that in any scientific field, there is always a large gap between the ultimate goal or aspiration and the resources and pathways actually available today?

Zhang Peng: That gap is exactly what drives an industry or a technology forward. It is like osmotic pressure in physics and chemistry: when the concentration differs on the two sides, it creates movement. The same is true here. When there is a gap between your ambitious goal and what your current resources and technologies can actually achieve, that gap stimulates people to keep researching, to find new methods, and to invest new resources. That is perfectly normal. The only question is how large the gap is.

Fang Sanwen: Is the equilibrium created by that osmotic pressure a kind of dynamic equilibrium?

Zhang Peng: Yes, absolutely. AI’s development has gone through repeated cycles of boom and bust. Why was there a second wave after the first winter? Because people once again saw the possibility of a new dynamic equilibrium. New resources came in, the “osmotic pressure” changed, and it seemed possible to overcome or make use of that gap. So people invested in new methods and new resources. It is always a dynamic process.

Fang Sanwen: In 1981, the first computer equipped with a GPU appeared. What is a GPU? It is very hot today — what is its relationship with AI?

Zhang Peng: A GPU is relative to a CPU. A CPU is the core processing unit of a computer, the central processor. A GPU is a graphics processing unit, originally designed to take some graphics-processing work off the CPU’s shoulders. The interesting thing is that GPUs are especially well suited to floating-point computation, whereas CPUs are more focused on integer computation. GPUs were specifically designed to strengthen floating-point capabilities. That happens to align very well with AI and scientific computing, both of which involve massive amounts of floating-point calculation. Some scientists therefore began asking whether hardware acceleration could speed up scientific-computing algorithms. Nvidia recognized this early. Jensen Huang reportedly sent GPU cards to many scientists with one simple request: run your algorithms on them and let us see the results, so that the company could demonstrate their value.

Fang Sanwen: Then in the late 1980s and early 1990s, AI entered a second winter. What happened in between, and how was it different from the first one?

Zhang Peng: The second upswing came because people found a new method: expert systems. Building on first-generation methods like the perceptron, researchers used structured knowledge representation to give computers expert-like specialized knowledge so they could answer questions. That was the success of second-generation AI. But after some development, people realized that even though this approach was theoretically more complete, there were still major practical problems. In principle, you could write down human rules in if/else form and hand them to the machine. But once you scaled up, could all knowledge really be exhaustively enumerated? In medicine, for example, could you fully enumerate all diseases and all treatment options? Writing all of that down might require an astronomical amount of effort. That is a problem of implementation cost and time. Second, even though computing power had improved, once you injected large amounts of expert knowledge, computation grew explosively. It still could not meet real needs. The “water level” is always changing: you see computing improve, so you introduce new methods and add more complexity, and then once again you discover computing power is insufficient. It is a dynamic, mutually reinforcing process.

Fang Sanwen: In 1997, Deep Blue defeated the world chess champion. What was the significance of that?

Zhang Peng: It was hugely symbolic. People tend to think conversation is easy, but chess is much more complex and implies a much higher level of intelligence. If a machine can outperform a human in chess, then it suggests machine intelligence has reached a certain threshold. That was the significance of Deep Blue defeating Kasparov.

Fang Sanwen: Almost 20 years later, in 2016, AlphaGo defeated Lee Sedol. How was AlphaGo technically different from Deep Blue?

Zhang Peng: The implementation path was different. Deep Blue mainly relied on search methods. Given known game records or the current board state, it searched for the next move, looked ahead a number of steps, and used Monte Carlo tree search and pruning. It did not search the entire space, only part of it, and tried to find the best possible solution within limited time. AlphaGo followed a similar overall path, but with a different implementation: it used neural networks and large amounts of data to approximate the search and prediction process. Both aimed to search the space and predict the best move, but they predicted in different ways. You can think of AlphaGo as more of an end-to-end approach, while Deep Blue was more rule-based, a pipeline with one component feeding into the next.

Fang Sanwen: In simple terms, what is deep learning?

Zhang Peng: Deep learning refers to algorithms with many layers. A single layer of neurons or computing units can solve only very simple problems. For more complex problems, you stack many layers on top of each other. The more layers there are, the more situations the model can represent, and the more complex the computation becomes.

Fang Sanwen: So it is basically functions wrapped inside functions?

Zhang Peng: Exactly. One layer nested inside another, built up continuously.

Fang Sanwen: From the perspective of end-user infrastructure, PCs spread around 1995, and around 2000 the internet became widespread, connecting machines and data. Was that related to the development of deep learning?

Zhang Peng: Absolutely. Deep learning requires a huge amount of computing power. That is why GPUs became so important: their floating-point capabilities are strong. Improvements in computing hardware were a prerequisite for deep learning’s rapid development. Compute is like an engine: the stronger it is, the more power it can produce. Data is like fuel: you need large amounts of good-quality fuel for the engine to run longer and produce more output. PCs and the internet provided vast quantities of data.

Fang Sanwen: Google published the Transformer paper in 2016, and not long afterward OpenAI released ChatGPT. Are those two things connected?

Zhang Peng: Definitely. OpenAI really pivoted toward large models after that paper was published. OpenAI was founded in 2015, and up to around 2018 it was not on this track — it was more focused on reinforcement learning. Not long after the Transformer paper came out, around 2018, scientists led by Ilya made a decisive shift and began GPT-related research based on that architecture.

Fang Sanwen: When the first version of ChatGPT came out, did you pay attention to it?

Zhang Peng: There were really two phases. When they started working on GPT in 2018, it did not attract that much attention and the results were not that strong. GPT-2 got some attention, and the academic world debated it with mixed views. Some said it was just brute force rather than an algorithmic breakthrough, while others saw it as a promising new paradigm. At that point there was more discussion overseas and relatively less in China. Thanks to the Tsinghua environment, we had more exposure to the international academic world. What really made people sit up and take notice was GPT-3 in 2020. Once GPT-3 came out, we focused on it and realized it might represent a turning point in the paradigm itself.

Fang Sanwen: Why did you make that judgment? What breakthroughs or phenomena did you see at the time?

Zhang Peng: Before that, we had been using traditional machine-learning methods for NLP tasks such as dialogue and question answering, but we had never really reached a satisfying result. Traditional NLP required a long algorithmic pipeline to process a sentence: identify its structure, nouns, verbs, and so on, and then try to understand it. But with GPT, we discovered you did not have to do all that. You could just feed the sentence in and it would answer — and answer very well. That was end-to-end problem-solving without the need for all that complex decomposition. In many cases, it crushed the traditional pipeline approach. That made people realize this method had clear advantages, and that it might be the next technological paradigm.

Fang Sanwen: The release of DeepSeek-R1 seemed to overturn a lot of people’s assumptions. In terms of both principles and performance, what breakthrough do you think it represented?

Zhang Peng: DeepSeek attracted a lot of attention in the industry, but it is still following the same broad path. It is not like GPT compared with traditional machine learning, where the methodology itself was fundamentally different. DeepSeek is not a methodological-level break. What it has done is more about reducing cost and optimizing engineering — pulling the industry back from the simple logic of “just throw in more parameters and more data.” It showed people that you do not necessarily have to keep scaling blindly; you can optimize the algorithms, reduce cost, and improve results at the same time. Even more importantly, at that moment it open-sourced the technology and handed it over to the community, academia, and industry for free use. That had a huge impact on the market.

Fang Sanwen: AI development involves compute, data, and algorithms. Which of these is the biggest bottleneck right now, and which kind of breakthrough is most likely to drive the next stage of development?

Zhang Peng: The answer changes over time. At first, people thought algorithms were the key and could take us directly to AGI. Later, when model parameter counts grew, people began to worry that the data available on the internet was not enough, and that pre-training might hit a wall. Then people worked on solving the data problem, and now that seems more manageable, so people worry that compute is insufficient. Later, technology advances and Nvidia keeps producing more chips, so compute becomes more available. Then people swing back and say the algorithm itself has problems — that efficiency is too low, catastrophic forgetting is unresolved, and maybe Transformer itself will eventually need to be replaced. So it is a dynamic cycle, an upward spiral.

Fang Sanwen: Do you see AGI as an abstract goal or a concrete one? Is it ultimately achievable, or can we only keep approaching it asymptotically?

Zhang Peng: AGI is not defined as clearly as AI. Its meaning and boundaries have never had a fully authoritative definition. But can AGI be achieved? As long as we define the goal properly — not in some absurd way — then in all likelihood it can be achieved. The real question is how long it will take. From the first day Zhipu was founded, our goal has been AGI, and we have our own definition of it.

Fang Sanwen: For the general public, does AGI feel a bit like the ultimate truth in science — something hard to make concrete?

Zhang Peng: Ultimate truth is hard to depict because nobody really knows what it looks like. But in science, there are different ways to define things. Turing, for example, defined intelligence behaviorally. AGI is often approached in the same way. One view is that once AI reaches the average human level across various capabilities, that counts as AGI. Once you have a behavioral definition, you can work backward: what technical capabilities do we need to match the human level? So the path depends on how you define the goal.

Fang Sanwen: In today’s public debates over AI, has anyone really gone beyond the nine objections Turing raised?

Zhang Peng: Basically not. AI has gone through several ups and downs over more than 70 years, but the questions keep coming back to the same place. The philosophical issues were identified very early on. What has changed is not the questions themselves, but the paths and methods we use to approach the ultimate goal that was defined from the beginning.

The boundaries of AI applications

Fang Sanwen: What exactly is a large model?

Zhang Peng: First of all, the word “model” is not hard to understand — it is basically the carrier of an algorithm. Traditional machine learning also had models, so there is nothing novel about that. The key is the word “large.” Why do we call it a large model? Going back to deep neural networks, you can think of a model as a giant computational matrix, where each element of the matrix is a parameter. Input data goes through matrix multiplication and addition to generate an output. That matrix is the core of the model. In a large model, the matrix is extremely large, which means the number of parameters is extremely large. A traditional perceptron might have only two or three parameters; a large model may have tens of millions, hundreds of millions, billions, or even hundreds of billions of parameters. That is why it is called a large model.

Fang Sanwen: Even non-specialists roughly know that there are different categories of large models. One is the general-purpose model, like ChatGPT, Grok, Gemini, or Zhipu’s GLM. Another consists of models tailored for specific domains or scenarios. Is that a valid two-part division?

Zhang Peng: After ChatGPT became popular, there was a lot of debate in China over whether it made sense to divide the field into general-purpose models and vertical or domain-specific models. We prefer to go back to first principles and ask: why would you divide them this way? We belong to the general-purpose model camp. At the time, some people argued that general-purpose models were trained on broad data, had huge parameter counts and high costs, but could not solve specialized problems. Others argued that you should build smaller models trained on domain-specific data to solve professional tasks — so-called vertical or specialized models. But later we found a paradox. If you already have access to specialized data, why not simply add that data into the training of the general-purpose model? Would it not then be able to solve the specialized problem too? Why build a separate specialized model? The second issue is whether a smaller domain-specific model trained on limited data will necessarily outperform a general-purpose model. In practice, once the general-purpose model incorporates specialized data, its performance often surpasses that of the specialized model. So the two fundamental premises behind specialized models start to collapse. In practice, people have increasingly seen this as something of a false proposition.

Fang Sanwen: That seems to point toward a rather alarming conclusion: all large-model companies are competing within the same general-purpose model business model. What will the competitive landscape look like in the end? Will there always be many general-purpose large models, or will the number keep shrinking?

Zhang Peng: Convergence is inevitable. Building these models is very expensive. It requires enormous amounts of compute, data, and talent. If everyone starts from scratch, then resources are being duplicated. From the standpoint of resource optimization, the field will naturally converge around a few leading players, while others will do different things. That is the broad direction. What we are seeing now is more specialization — not only in applications, but also in model infrastructure, platforms, specific vertical use cases, and services. That is how a broad ecosystem forms. But the barrier to entry for the base model itself is simply too high for most companies.

Fang Sanwen: The world’s main players in general-purpose large models are probably four or five, mainly in China and the US, with Europe’s Mistral as another example. Do you think it could eventually come down to just one?

Zhang Peng: My intuition is no. Technological development needs diversity, especially in the early and middle stages, when there are many possible directions for innovation and research. Different players have room to differentiate and survive. On top of that, the market is huge and growing quickly, so there is enough space. It will not converge overnight. In the early and middle phases, you will have a diverse ecosystem. It will not become winner-take-all just because one company gains an advantage. A single dominant player becomes more likely only when the technology becomes relatively stable and innovation slows down, because then you get stronger Matthew effects. At the current stage, that is not what we are seeing.

Fang Sanwen: What mainly determines the differences between large-model companies — compute, data, or algorithms?

Zhang Peng: All of them matter. Compute reflects resource investment: what kind of compute you can get, in what quantity, and at what cost all determine the efficiency, speed, and capacity for innovation. Data is similar: if you can access high-quality data in certain areas, you gain an advantage there. But algorithms are even more fundamental. Whoever can innovate faster and more consistently at the algorithmic level will rank higher in the industry.

Fang Sanwen: Some people think that for ordinary people, AI has mostly just made it easier to retrieve and organize information, without yet changing daily life that much. Do you think that broader change will come?

Zhang Peng: It definitely will, and in fact it is already happening. In office work, jobs, and daily-life scenarios, AI can already help us do many things. I am not someone who enjoys wandering around shopping malls. For many tasks, I would much rather have AI handle everything with one click. That is why we launched AutoGLM. On my phone, if I want to buy something on an e-commerce platform, I may just have an idea and ask AI to pick out a few options, put them into the shopping cart, and optimize for the lowest price or the best value for money. I would only need to confirm and pay at the end. Things like that are already starting to happen in everyday life.

Fang Sanwen: Can you give some examples of industries that AI has already changed?

Zhang Peng: In some areas, the changes are not very visible to the public — take industry or pharmaceuticals. Traditionally, developing a new drug took a very long time. You had to screen huge numbers of compounds, run experiments, and spend a lot of money. Now AI can help with drug design and molecular screening. For example, when you are trying to match small molecules to protein structures, AI can do much of that large-scale screening work. Then there is AlphaFold. In the past, figuring out protein structures required painstaking experimental reconstruction. Now AlphaFold can use historical data to generate predictions much more quickly, allowing researchers to identify promising candidates computationally before confirming them in the lab. That has already changed those fields enormously, and in time it will affect ordinary people as well — maybe drugs will not be so expensive in the future.

Fang Sanwen: Can AI help doctors, or even replace them in some roles?

Zhang Peng: Definitely. Both in China and abroad. There is a large body of foreign research literature and clinical data that AI can analyze to help doctors handle difficult cases or support medical research. In China as well, hospitals and commercial firms are developing related products to help primary-care doctors with knowledge support, training, and so forth.

Fang Sanwen: Is assisted driving or autonomous driving also an AI application direction?

Zhang Peng: Autonomous driving has been worked on for more than a decade. Replacing human drivers may be possible, but like AGI, the first question is how we define the essence of “full autonomy.” What problem exactly are we trying to solve? Once you describe the goal clearly, then you can ask what current methods can achieve, what their defects are, and what the next methods need to be. It all depends on the definition. We can only keep approaching the part we understand through behavior, making the system look more and more like the target behaviorally. But because we still have not fully unpacked the essence of intelligence or how it forms, we cannot guarantee 100 percent success.

Fang Sanwen: So if it is a kind of imitation game, the path is simply to keep getting closer?

Zhang Peng: Exactly. Keep approaching it.

Fang Sanwen: But if you had fully cracked it — understood the whole inferential process — then in principle you could reproduce it without limit?

Zhang Peng: That is the difference between a black box and a white box. In AI’s evolution, the first and second generations were more like white-box approaches. But people found that path very exhausting. Then with third-generation deep learning, the field gradually shifted toward black-box approaches, because white-box approaches ran into too many unresolved problems. Some people said: forget about explaining every internal mechanism — the human brain is also a black box. Just use a black box to model a black box, and judge it by the input-output results. The outcomes turned out to be surprisingly good, so people shifted to end-to-end methods.

Fang Sanwen: Even if a black box can get extremely close, that does not necessarily mean it can reproduce the target with unlimited precision, right?

Zhang Peng: There are two sides to that. First, if your black-box model gets arbitrarily close to a human, then of course it is reproducible in the sense that it is just data — you can copy it. But what remains non-reproducible is that you cannot fully decompose it and say exactly why it works, or isolate certain capacities and reconstruct them in a white-box way. So the issue has two sides.

Fang Sanwen: What exactly is the relationship between humans and AI? A lot of people say: the more capable AI becomes, the more likely I am to lose my job. How do you see it?

Zhang Peng: It is complicated, and it is something everyone has to confront. This generation of AI is different from previous ones because it can finally engage in something close to an equal conversation with humans, and in terms of knowledge it may exceed the average person. That creates a real challenge. Humans only know how to deal with other humans; we have not really had to coexist with AI before. We do not yet know how to interact with it harmoniously. But I do not think human intelligence is going to stop evolving. Humans are extremely adaptable. Throughout history, whenever there were technological or social upheavals, people worried that humanity faced some huge crisis or would be replaced — but we got through it, and in many ways life kept getting better. Human beings themselves are also advancing and evolving. So AI may not be an absolute crisis; it may even accelerate human evolution.

Fang Sanwen: What is the biggest difference between human intelligence and machine intelligence?

Zhang Peng: We once divided the AGI path into L1 through L5: knowledge learning and compression, reasoning, self-learning, rudimentary consciousness, and full consciousness. Large models today have probably progressed to somewhere around the middle stage of self-learning. The real difference comes in the latter two stages: humans have self-awareness. We know that “I am me.” AI obviously cannot do that yet.

Fang Sanwen: The biggest AI competitors globally are the US and China. In your view, what exactly are they competing over?

Zhang Peng: Personally, I think it is a competition between two different development paths or philosophies of AI.

Fang Sanwen: What is the American path, and what is the Chinese path?

Zhang Peng: The US seeks extreme innovation in AI. It is all about reaching the highest frontier through concentrated capital and concentrated effort among a small number of top players. China is different. It is much harder in China to concentrate such massive resources on a small set of firms and just brute-force the outcome. In China, there is much more emphasis on certainty and broad accessibility. So the Chinese path is one of steady, grounded progress: first, innovation has to keep up, and we cannot lag too far behind; second, at every stage, the goal is to convert technological progress into productivity and economic value that improves life. Policymakers in China talk a lot about “AI plus all industries,” AI empowerment, and improving livelihoods and the economy. China’s approach is not to fly past everything and focus only on the final destination; it is to keep laying eggs along the way, making each stage of progress useful across different sectors. That path is much more efficiency-driven and ROI-driven, with a stronger emphasis on cost-benefit calculations.

Fang Sanwen: So as in many industries before, China focuses more on implementation, applications, industrialization, and commercialization, emphasizing user experience and commercial efficiency. Do you think AI competition will continue to follow this division of labor for quite a long time?

Zhang Peng: Possibly, yes. At least in our country, that will likely continue, because it reflects deeper historical and cultural factors. The top US firms are focused on going from zero to one. China is more focused on going from ten to one hundred — and in fact on the steps from one to ten and ten to one hundred — because that is how you achieve broad accessibility at scale.

Zhipu’s business model and moat

Fang Sanwen: Zhipu is itself a case of AI deployment, industrialization, and commercialization. Could you briefly explain Zhipu’s business model?

Zhang Peng: Zhipu has thought about this quite clearly. Around 2020 or 2021, we were already working through this issue. The technology did not originate with us, but we caught up quickly. At that time, we asked ourselves how this technology could become a business — what the commercialization path would be. We proposed MaaS, Model as a Service: turning the model itself into a service that people can understand, use, and embed into their products, systems, and daily lives. That became our business model.

Fang Sanwen: Was that the model you settled on at the company’s founding, or did it emerge gradually?

Zhang Peng: It emerged gradually. In the early days after the company was founded in 2019, we were still focused more on academic applications and exploring some services. Later, an independent external team explored MaaS commercialization and served more categories of clients. At a certain point, we concluded that this model was the right one, so we integrated that team back in. Now it has grown quite a lot and is performing very well, which also shows that MaaS is currently one of the more viable commercialization paths for large models.

Fang Sanwen: Is this model the best temporary answer, or is it a stable long-term form?

Zhang Peng: I think MaaS will remain relatively stable for quite a long time, but it is definitely not the end state. There are still many uncertainties about the eventual destination. One thing that seems clear is that large models are increasingly evolving into infrastructure — something like water, electricity, or gas: intelligent infrastructure that society needs to function. Infrastructure has to be standardized, affordable, and easy to access. MaaS fits that shape very well. Another direction is that beyond the model as infrastructure, there will also be very different forms of applications, possibly integrated with hardware and embedded in phones or terminal devices to create new kinds of products. It is like electricity and electrical appliances — there are big market opportunities at both ends.

Fang Sanwen: So on one side there are base models, and on the other side applications built on top of them. Are you more focused on the latter?

Zhang Peng: Not exactly. We do both at the same time. But our main focus is still the foundation-model side of MaaS. We do also have some applications on top.

Fang Sanwen: Could you give a concrete example? What kind of MaaS service do you provide to your clients, and what problems does it solve?

Zhang Peng: There are many examples. Our major clients include nine of China’s top ten internet companies. One client experienced a situation last year where an international event caused a large number of users from an overseas social-media platform to migrate to a domestic social platform, but there were language barriers: foreign users could not understand Chinese, and Chinese users could not understand foreign languages. We used our model to help solve a large volume of translation work. Another example is our cooperation with Samsung. We integrated model capabilities into the handset itself, so users can use them on-device, which helps address data-privacy concerns. Chat records and images do not have to be uploaded to the cloud; they can be searched and edited locally. Those are all examples of solving real-world problems.

Fang Sanwen: How large do you think the AI applications market is?

Zhang Peng: Gartner estimates it will be in the trillion-dollar range. Some reports say the global AI market could reach $4.8 trillion by 2033. In China alone, it should be at least in the trillion-renminbi range.

Fang Sanwen: Who are your main competitors at the moment?

Zhang Peng: We are an independent general-purpose large-model provider, and there are not many players of that kind in China. But the big tech firms are also doing similar things — building base models and AI-related businesses — so they are our peers, competitors, and partners all at once.

Fang Sanwen: Relative to those competitors, what is your company’s enduring advantage?

Zhang Peng: We are highly focused. First, our understanding of AI has been ahead of the market average from day one. Second, we positioned ourselves from the beginning as a general-purpose large-model company with AGI as the goal, and we are willing to not do other things in order to stay focused on that. That is our biggest advantage. Based on that positioning and concentration, we can create market opportunities through technological innovation and product iteration, and then turn those into commercialization opportunities. The capabilities of a general model are not something abstract; at every stage they can be translated into concrete applications. Recently we have been particularly focused on coding ability. That is a general-purpose capability, and once you focus on it, the technical, product, and commercial value can be enormous. That comes from understanding the importance of the issue early and accurately, and from having top-level technical capability.

Fang Sanwen: Is your advantage mainly in the model itself, or in deployment and applications?

Zhang Peng: You cannot separate the two. In this AI wave, the loop from algorithm research to engineering implementation, productization, and user feedback has become extremely compressed. It is not like the old days, when a lab would publish a paper and years later someone might turn it into a demo, then a product, then iterate again. In this wave, it took only about five years from algorithmic innovation to something like ChatGPT going online and then instantly reaching hundreds of millions of users. The cycle has been compressed dramatically. So you cannot say, “We will perfect the algorithm first and think about delivery later.” That is no longer possible. It all has to be integrated. We research while simultaneously shipping, so people can use the product, give feedback, and help us make it more useful.

Fang Sanwen: There may be fewer and fewer general-purpose large-model providers, but will the number of players taking model capabilities into specific enterprise scenarios keep growing?

Zhang Peng: Absolutely. That “last mile” of turning model capabilities into customer needs and product features will definitely keep expanding, because demand is huge. That is the direction the ecosystem is heading.

Fang Sanwen: If many players enter that space, could profit margins get driven very low? For example, image-recognition capability was eventually deployed into many different scenarios by many vendors, and margins became thin.

Zhang Peng: That kind of thing tends to happen when the technology approaches its ceiling. For example, once face recognition reaches 97 or 98 percent accuracy, pushing a little higher may not matter much. The technology becomes stable, everyone piles in, they compete on cost, and prices get driven down. But large models are still in a phase of rapid technological growth. The technology premium is still high, and innovation is still very active. So we are not yet at the stage of pure low-price competition. What we need to do is stay ahead in innovation while the curve is still rising, and use that pace of innovation to create market space and capture the premium that comes with it.

Fang Sanwen: Is the main purpose of your R&D spending to maintain a leading edge in large models themselves, or to improve productization and commercialization in terms of user experience and efficiency?

Zhang Peng: Definitely the former. The upper limit of our base model capability is the foundation of everything. All commercialization rests on that.

Fang Sanwen: This industry includes both domestic and international giants. Is it difficult to maintain an advantage?

Zhang Peng: There are definitely challenges. But precisely because there are challenges, the team feels we have to do it — and we have to succeed. We have that confidence.

Fang Sanwen: The major platforms can invest far more money than you can, so it is hard to compete with them on compute; in terms of data, they also have huge reservoirs of internet content. Does that mean your edge lies in algorithms?

Zhang Peng: The three elements should not be viewed in isolation. On compute investment, of course we cannot match the big platforms — they can afford much more. On data, I would also not assume they can simply take all those data resources and legally use them for training for free. There are legal issues involved. Algorithms and R&D capability are our strengths. If you look at the three elements separately, everyone has strengths and weaknesses. But how you combine them and produce a real chemical reaction — that tests the capability of the team. And even more fundamentally, it comes down to how deeply you understand AGI and AI at the level of first principles. Let me give one example: a major platform also began working on large models relatively early, but after a period of investment, its team was asked how to commercialize it, was pushed hard into commercialization, failed, and was then replaced. The big firms do not have unlimited patience either; they have performance requirements. So there is no need to overstate the supposed omnipotence of the big platforms.

Fang Sanwen: It sounds like this is a market that takes courage to enter. Are you a courageous person?

Zhang Peng: Our team has a lot of courage.

Fang Sanwen: Your business model is built mainly around the enterprise market. What are you doing on the consumer side?

Zhang Peng: We thought about that very early. Why should we even divide things into B2B and B2C? What is the principle behind that distinction? Nobody can explain it clearly. At the commercial level, people say B2B and B2C products differ in form and payment model. But from the perspective of product and technology, if you go back to first principles, there is no essential difference, because whether you are serving enterprises or internet users, in the end you are serving people. People will only pay if they recognize the value and get real gains from the technology and the product. The difference lies in the payment logic and the decision-making logic of B-side and C-side customers. But AI is productivity at its core, and payment depends on the value added through gains in productivity. Without that, nobody pays. So the consumer side is not our current focus. It is like electricity: whether the user is an individual or a company, what is the real difference? There is none. In both cases, people pay because it is useful.

Fang Sanwen: So for now, commercialization is still mainly on the enterprise side?

Zhang Peng: The future is highly uncertain, which is why I keep saying that the biggest challenge in this era comes from human beings themselves. When faced with something new, people can only linearly extrapolate from the past to predict the future, and they can never predict things that lie outside their existing frame of reference. A lot of problems ultimately come from that. Zhipu should keep moving forward with the AGI ideal. When the market needs a certain form, we will become that form. At this stage, we believe this model is a good fit for us, so we will keep moving down this path. As for the much more distant future, we cannot predict it, and there is no need to artificially limit ourselves.

This marks the second time a Chinese large-model research team has published work in Nature following DeepSeek’s cover feature, and it is also the first time a domestic research institute in China has had a paper published in the journal’s main edition.

Most current multimodal models rely on separate processing pathways for text, images, and video, leaving open the question of whether a single unified approach is possible.

A recent Nature study from the Beijing Academy of Artificial Intelligence (BAAI) suggests that autoregressive modeling—the same next-token prediction paradigm used in large language models—may provide such a unifying framework.

Built on the multimodal model Emu3, the research shows that a purely autoregressive architecture can handle both perception and generation tasks at performance levels comparable to specialized state-of-the-art models. The unified framework also extends naturally to applications such as robotic manipulation and multimodal interactive content generation.

In an interview with The Intellectual, Huang Tiejun, chairman of BAAI and professor at Peking University, explained how Emu3 enables multimodal unification and discussed the broader technical pathway toward AGI.

The Path to AGI Has Been Found — The Next Step Is to Fully Explore It

The Intellectual:
AI capabilities have advanced rapidly in recent years. BAAI has been closely following developments in the field. Looking back at the breakthroughs of the past few years, what do you think was the truly decisive turning point?

Huang Tiejun:
From 2018 to the present, people have discovered a technological path that actually works: the autoregressive route. This approach is based on the Transformer architecture and trains models by predicting the next token in a sequence. That was the most important breakthrough from zero to one, and this path leads toward artificial general intelligence (AGI).

BAAI has always held a conviction: if the combination of Transformers and next-token prediction has completely worked for language models, can it be extended to all modalities of data—language, images, video, and even multimodal data such as vision–language–action (VLA)? From a methodological perspective, I believe this is entirely feasible.

When people talk about language, images, and video today, they are really referring only to the most common and easily understood forms of data. In fact, this method can accommodate any type of data, including data from different layers of the real world.

However, this is still just our belief. To truly realize it, we must continue testing it with real data. Technological innovation can only be filtered by time and proven by results.

The Intellectual:
You see 2018 as a turning point. What changed around that time?

Huang Tiejun:
Before 2018, artificial intelligence was mainly human-designed intelligence. Humans designed the intelligence. Whether knowledge bases or expert systems, designers controlled every piece of logic behind the system almost like gods. This reflected a traditional scientific mindset: first fully understand the underlying principles, and then manually design a system based on those principles.

After 2018, with the birth of the first generation of GPT, so-called generative AI emerged, and the methodology fundamentally changed. Many people interpret “generation” as the system producing text, images, or videos. But I prefer to interpret generation in a way similar to the emergence of life on Earth—namely evolutionary generation.

On Earth, life evolved from non-life, and from simple forms to complex ones. Is there an incentive mechanism behind that process? Of course there is. But we still do not clearly understand what that mechanism is. Life science and brain science have been studying these questions for many years, yet overall they remain a kind of “dark forest.” We are only gradually discovering the principles behind them.

The changes after 2018 are similar. People have found a feasible technological route toward general artificial intelligence: training models through data-driven methods so that intelligence emerges. Yet the interactions that occur within this technological route remain unclear.

The Intellectual:
You said the autoregressive route is the only path toward AGI. But there is much debate about the definition of AGI. What is your view?

Huang Tiejun:
My view is that artificial general intelligence has already been realized to a certain extent.

According to the traditional way of thinking, people feel it has not been realized yet, because the underlying principles have not been fully understood. How can we say it has been achieved if we do not understand how it works? But today’s large models already exhibit strong general capabilities. You can test them. In terms of capability, they are stronger than many humans. Under such circumstances, insisting that they are not general intelligent systems becomes somewhat unreasonable.

Changes in people’s understanding of AGI are also related to changes in AI concepts throughout history. The earliest definition of general intelligence focused on behavior, function, and performance—essentially the Turing Test. If a third-party evaluator cannot distinguish between a human and a machine during interaction, then the machine has passed the test. Today’s large models have already reached this level.

The term AGI emerged roughly in the late 1990s, only a little over twenty years ago. People generally interpret AGI simply as general artificial intelligence. But strictly speaking, the concept proposed in the 1990s was actually harder to realize—it assumed AI would need self-awareness.

If AGI means AI with self-awareness, then I believe it has not yet been achieved, or at least it remains an open question. But if we do not adopt such an overly strict definition, and instead define AGI as a system capable of performing a wide variety of tasks like humans—possessing generality—then I believe we already have it.

The Intellectual:
Why can the autoregressive route bring about such transformation?

Huang Tiejun:
This method captures the essence of how intelligence evolves. “Predicting the next token” seems simple, but it actually touches the core problem of intelligence. Because every intelligent system essentially does one thing: it uses the past to infer the future.

Animals rely on past experience to decide whether to flee. Humans infer economic trends from historical data. We read books in order to improve our ability to judge the future. The most basic function of intelligence is to increase the probability of making reasonable predictions in uncertain environments. The evolution of biological intelligence is essentially a process of increasing the probability of making correct choices.

This path contains two indispensable components. The first is the Transformer. If we draw an analogy with life sciences, it represents the “structural foundation.” In biology we say “structure determines function.” The kind of DNA an organism has determines its physiological form. In the AGI domain, the Transformer plays the role of that fundamental structure.

But structure alone is not enough. Intelligence evolves gradually through interaction with the surrounding world. The human brain works the same way: intelligence is not formed all at once but evolves as the environment changes. This is what we call function shaping structure—the pressure of the environment drives structural changes.

In artificial intelligence, this evolutionary process depends on data-driven learning. Large models learn patterns through autoregressive training, meaning they repeatedly predict the next token. Each prediction is an attempt. If the prediction is wrong, the model adjusts its internal parameters according to the data. If it is correct, those connections are reinforced. Under the influence of massive data, the model gradually learns the patterns of language, logic, and even multimodal information. The combination of Transformers and autoregressive training satisfies the fundamental conditions for the evolution of intelligence.

The Intellectual:
How exactly does next-token prediction work?

Huang Tiejun:
Tokens are the basic units of natural language processing. They can be words, phrases, or roots; they can also be punctuation or artificially defined markers. Essentially they are symbols. There are two ways to understand the meaning of symbols. One is direct sensory experience. But AI has no body, so it can only learn meaning through relationships among symbols.

Before 2018, early word-embedding methods used statistical co-occurrence relationships between words to map each word into a high-dimensional vector space. Words that frequently appeared together were placed closer together in this space.

But those representations were static. No matter what context a word appeared in, its vector representation remained largely unchanged. The model learned the average meaning of a word rather than its dynamic role in a particular context. In other words, it solved the question “what does this word mean?” but not “what does this word mean in this sentence?”

The emergence of Transformers changed this. Take the novel Dream of the Red Chamber as an example: clues and foreshadowing across dozens of chapters are interconnected. Understanding a character requires not only knowing their name but also examining their interactions and experiences. What Transformers can do is discover relationships between any two tokens within a given sequence. In the context of Dream of the Red Chamber, it means computing the correlations between any two characters in the entire book.

Human intelligence also works this way when reading a novel or long text: it builds relationships within context and repeatedly analyzes the logic. The model is essentially doing the same thing, only at a much larger scale and in much higher dimensions. It not only understands the content, but in many cases understands it more thoroughly than most human readers.

Therefore, when the model predicts the next token, it is not simply performing frequency statistics. It is invoking a highly complex structure that compresses the relationships within the entire context. Prediction is merely the surface manifestation; what actually happens is that the structure internalizes patterns and then infers future developments through those relationships.

Let AI Predict the Physical World the Way It Predicts Language

The Intellectual:
AI systems already demonstrate impressive capabilities. But many researchers argue that unless we can fully explain the internal mechanisms of these models, they cannot truly be considered general intelligence.

Huang Tiejun:
To be honest, that is a typical bookish mindset. After DeepSeek caused a global sensation, DeepMind CEO Demis Hassabis commented that “DeepSeek may be China’s best AI model, but it has not demonstrated any new scientific progress.” That kind of criticism looks at technological innovation through the lens of traditional science.

If we must draw an analogy, many major technological breakthroughs in human history followed the pattern of “technology first, scientific theory later.” When the Wright brothers built the airplane, aerodynamics was far from mature. The principles of flight had not yet been fully explained by theory. Yet airplanes still flew and changed the world.

AI development today is at a similar stage. Large models are closer to engineering innovations than to pure scientific exploration in the traditional sense. Through the methodology of “predicting the next token,” humans have already created intelligent systems with general capabilities. That practical success is undeniable.

Another point must be clear: intelligence itself is extremely complex. It cannot be reduced to a few rules or formulas. Simply because something does not conform to a particular theoretical framework does not mean we should deny the intelligence demonstrated by current models. That would be as absurd as refusing to acknowledge that airplanes can fly.

The Intellectual:
But if we never understand the principles behind large models, can this kind of technological innovation be considered rigorous science?

Huang Tiejun:
Understanding the principles is not necessary. When I say “not necessary,” I do not mean that principles are useless or undesirable. I mean they should not be a prerequisite. Once people start emphasizing necessity, many assume that we must first invent a theoretical framework before moving forward. I believe that mindset actually limits people’s ability to make larger contributions, because their thinking becomes too rigid. I used to think that way myself, but eventually I liberated myself from it.

We have discovered an effective methodology that can convert massive amounts of data into intelligence. That method works. The mechanisms behind it are questions for future scientific research, but they should not become a reason to stop technological innovation. We should not deny objective technological results simply because they do not fit within familiar scientific frameworks.

The priority now is engineering and scaling—pushing this path deeper and further. As for the scientific principles of artificial intelligence, future researchers will eventually uncover them.

The Intellectual:
If intelligence cannot be summarized into a few formulas, can we still establish benchmarks to measure its development?

Huang Tiejun:
We can set measurement metrics, but as intelligence becomes more complex, the measuring tools must also evolve. True intelligence has infinite complexity. We cannot force it to fit within static standards. Any finite measurement only provides a small window into understanding it, not the full picture.

The Intellectual:
You have repeatedly emphasized that large models are primarily a technological innovation. Yet top journals like Nature usually prioritize fundamental scientific contributions. Why did BAAI decide to submit the Emu3 research there?

Huang Tiejun:
I hope to correct the biases of traditional natural science. Many people trained in natural sciences are constrained by their own ways of thinking. They are accustomed to studying objects that already exist in nature and discovering the laws behind them.

Artificial intelligence is different. AI systems do not exist naturally; they must be created. They are technological inventions. In this sense, AI research is almost the opposite of traditional natural science. Applying the same mode of thinking from one direction to the other is fundamentally misguided.

Many people keep asking, “What are the laws behind artificial intelligence?” But laws can only be studied after the object itself exists. Life exists, so we can study the laws of life. But artificial intelligence systems are still being built. If we demand a complete theory before building them, we effectively block technological innovation.

If we wait until everything is theoretically understood before starting, we might not build AI even in 300 years. The history of technology never works that way. The normal pattern is technological breakthroughs first, scientific explanations later. First airplanes, then aerodynamics. First build artificial intelligence, then study the science of artificial intelligence. As the old saying goes: “When you understand what should come first and what should come later, you are close to the right path.” If the sequence itself is confused, using the standards of natural science to judge an entirely different direction is hardly something to be proud of.

Unifying Multimodality Through Autoregression

The Intellectual:
Your Nature paper argues that multimodal learning can be unified through the autoregressive approach. What do you see as the limitations of current mainstream multimodal models?

Huang Tiejun:
When people talk about multimodality today, they often think of “multiple modalities”—simply combining vision, audio, and text together.

For example, Transformers perform very well in text tasks but were not originally designed for multimodal problems. Image and video generation now largely rely on diffusion models, which generate high-resolution outputs through iterative denoising. For vision-language perception, many approaches combine CLIP encoders with large language models.

If the goal is to solve a specific problem within a single modality, designing specialized architectures can work quite well. But if every modality requires special patches and separate architectures, that cannot be called general intelligence. The real question is whether there exists a general route capable of handling intelligence across all modalities and all forms of data.

That is the value of the autoregressive route. It is also why we believe it is the core pathway toward AGI. Emu3 was developed under this philosophy. Our experiments show that even without diffusion models or hybrid architectures, a purely autoregressive model can achieve flagship-level performance in both perception and generation tasks.

The Intellectual:
Your paper mentions that Emu3 generates video purely through an autoregressive approach and performs comparably to diffusion models. What is the fundamental difference between these two approaches?

Huang Tiejun:
Diffusion generates content itself, but it is not the evolutionary generation I mentioned earlier. The two are fundamentally different.

The autoregressive approach is suitable for all types of data. By predicting the next token, it can model any form of information—images, videos, even robot actions. That is why we insist on the autoregressive route. It has strong potential to unify all modalities.

Diffusion models are excellent for generating images and videos. Their core idea simulates a physical diffusion process—like ink dispersing in water. Starting from noise and reversing the diffusion process produces an image or video. This method excels at generating visually realistic outputs, but it does not focus on the underlying relationships between objects in the scene. It is suitable for the relatively narrow domain of image generation.

When dealing with language or other abstract data, the situation changes. In language, words form complex semantic and structural relationships. Characters, events, and concepts in a novel form a massive interconnected network. This complexity far exceeds the interactions among molecules or pixels in the physical world. Diffusion methods cannot effectively model such relationships. They cannot capture deep logical connections between words or infer future developments.

The Intellectual:
Will future research extend this approach to other modalities?

Huang Tiejun:
The answer is already in the paper. We converted Emu3 into a vision-language-action (VLA) model and directly tested it on robotic manipulation tasks. On the CALVIN long-horizon benchmark, this general approach performs just as well as models specifically designed for robotics.

Another important point is that we directly use discrete encodings for vision, language, and action. Some other approaches require additional video training stages. This once again proves that autoregression is a universal logic. It does not require task-specific patches. Once the logic works, it naturally extends from perception and generation to embodied intelligence.

The work published in Nature was completed in 2024 based on the initial version of Emu3. By 2025 we released Emu3.5.

With this new version, we made a deeper discovery: as model parameters, data, and compute scale increase, the model begins to exhibit emergent abilities to understand and predict the dynamics of the physical world—spatiotemporal relationships and causal logic. This suggests that scaling laws do not apply only to language. When extended to the real world, which is even more complex and governed by physical laws, the same path still works.

The Intellectual:
Although Emu3 shows the potential of the autoregressive approach in multimodal settings, this is still an experimental path. What is still missing for building a true “world model”?

Huang Tiejun:
Recently many people have argued that scaling laws are reaching their limits. I believe that is incorrect. What has reached its limit is simply language-related data.

People talk about “world models,” but what exactly is the “world”? For a robot, does recognizing the world mean entering a room without hitting the table or grasping a cup with the right force? That is far from sufficient. The real world contains complex physical interactions. When you run into a wall, is it made of concrete or wood? If it is glass, can you pass through it? These kinds of physical and material properties are largely absent from today’s model training.

If we go deeper, interactions between atoms and molecules, or the hardness of concrete after it solidifies—are these not also part of the world? If they are, then scientific experimental data and molecular measurement data should also be used for training. Relying solely on the limited language and image data on the internet is not enough to support true general intelligence.

Even if we model everything humans currently know, we still have not exhausted the complexity of the world. The objective world is infinitely complex, and we can only continuously approach it. As long as that infinite complexity exists, and as long as we can introduce deeper scientific data, scaling laws will not reach an end.

The Intellectual:
BAAI has long supported scholars from both academia and industry. As an independent research institution, how does its work differ from research in universities or companies?

Huang Tiejun:
There are things that universities cannot do—not because they lack the capability, but because the necessary conditions are not in place. Building a systematic, operational project requires a team, funding, and time. In universities, professors can explore theoretical questions on their own, but developing a complete system requires first securing funding and assembling a team, which can take a long time. The pace of AI development simply does not wait for you to slowly go through the funding process. By the time you spend a year securing funding, the technological direction may already have shifted.

As for companies, they tend to be pragmatic. When a technical path has not yet been fully validated and remains more of a belief than a proven solution, companies are generally unwilling to invest heavily in trial and error. What businesses prefer is to take approaches that others have already tested and proven effective, and then quickly turn them into predictable products.

BAAI sits somewhere between universities and companies. We have relatively stable funding and teams. Once we reach a consensus that the autoregressive approach is the only universal path capable of handling all modalities, we simply move forward and build it. In engineering and technology, whether something works cannot be determined by persuasion—it has to be demonstrated through real results.

What we need to do is spend the time to build it. Once the path is proven to work, companies will naturally follow and invest much larger amounts of money to industrialize it.

Human Primacy Must Be Upheld in Military Applications of AI to Avoid Technological Runaway

人工智能军事应用应坚持由人主导、防止失控

Question: According to reports, the U.S. Department of War has demanded American technology companies to allow unrestricted use of AI technologies by the military. In addition, the U.S. military has employed AI tools extensively in military operations against Venezuela and Iran, triggering deep concerns over war ethics and the boundary of military application of technology. What’s your comment on this?

记者：据报道，美国防部要求美科创公司允许军方不受限制地使用人工智能技术。在针对委内瑞拉、伊朗的军事行动中，美军都大量使用人工智能工具，引发对战争伦理与军事技术应用边界的深刻担忧。请问对此有何评论？

Jiang Bin: Such choices as the unrestricted application of AI by the military, using AI as a tool to violate the sovereignty of other nations, allowing AI to excessively affect war decisions, and giving algorithms the power to determine life and death, not only erode ethical restraints and accountability in wars, but also risk technological runaway. A dystopia depicted in the American film The Terminator could one day come true. Upholding a people-centered approach and the principle of AI for good, China believes that human primacy must be upheld in military applications of AI, and that all relevant weapon systems must be put under human control. We are opposed to taking advantage of the lead in AI and other emerging technologies to pursue absolute military dominance, or undermine the sovereignty and territorial security of other countries. China will work with other nations to advance multilateral AI governance with UN centrality, strengthen risk prevention and control, and ensure that AI always develops in a direction conducive to the progress of human civilization.

蒋斌：不加限制地推进人工智能军事化，将人工智能作为侵犯他国主权的工具，让人工智能过度影响战争决策，让算法掌控人的生杀大权，不仅侵蚀战争的伦理与责任约束，还可能导致技术失控，甚至使美国电影《终结者》里的灾难场景成为现实。中方始终坚持“以人为本、智能向善”原则，始终认为人工智能军事应用应坚持由人主导，反对利用人工智能等新兴技术优势谋求绝对军事霸权、损害他国主权和领土安全。我们愿与世界各国一道，推进以联合国为核心的人工智能多边治理进程，加强风险预防和管控，确保人工智能始终朝着有利于人类文明进步的方向发展。

By contrast, official public statements from China tend to be more cautious and much less focused on explicit conceptual discussions of AGI. From the American perspective, this asymmetry in the intensity of discussion creates an information vacuum. In that vacuum, people often begin to speculate, fill in the gaps with their own assumptions, and sometimes construct exaggerated narratives—occasionally even wondering whether China might be hiding something akin to a “Manhattan Project–style” initiative.

Within China itself, the situation is also more complex than it might appear from the outside. Both academics and industry figures have discussed AGI publicly, but their understandings of what AGI actually means often differ significantly. Even when companies make public statements about AGI, those remarks frequently reflect personal views rather than a unified institutional position, because there is often no single, shared definition of AGI within the companies themselves.

Today, Song-Chun Zhu (朱松纯), Dean of the Beijing Institute for General Artificial Intelligence, had a one-on-one conversation with CGTN host Tian Wei and discussed his views on AGI.

Zhu Song-Chun is a highly influential Chinese scientist in the fields of computer vision, artificial intelligence theory, and cognitive science. He holds an important position in both the Chinese and American academic communities: he spent many years teaching AI at leading universities in the United States, and in recent years he has also become one of the key figures promoting the development of major AI research institutions in China.

According to Zhu, most people just use AGI as a term for fundraising to attract public attention. AGI is an agent that learns to grow, step by step, in mathematical space. His creation "Tong Tong" proves it. Starting as a "baby" recognising toys, she now handles household chores and picks up clues in certain situations as a "digital citizen."

Below is Zhu’s remarks:

It has been simplified as the United States is trying to reach for the stars by developing AGI, while China is more about AI plus. I think that’s how the media frame it. You don’t mask, and you wanna go to, let’s think about human survival, to go to Mars and so on. Sounds like they’re exploring the frontiers. But I tell you, they’re not, most of the people who talk about AGI right now from companies haven’t really done deep research about AGI. They’re just entrepreneurs who try to use AGI as a term for fundraising, to attract public attention. So I think that’s wrong direction.

But let’s come back to AGI. AGI is a big mathematical space. Tongtong is really the first AGI agent. At the beginning, we just say, okay, we start with a little room. It’s like when you have a cradle, right? Baby comes through the world. It’s very little room. And there were only a desktop with different toys I remembered around. And that’s the entire world for her. And then she began to develop certain so-called embodied physics tasks. Right now, people think about a physical model, right? And then the physics try to pick up objects and move them around, and then pile them up and form a certain shape, and so on. And then we enlarge the scope to a room she can move around, and then she can walk and dance, just somersault, just body movement. We further give her an apartment with three bedrooms, one guest room, a toilet, and a washing machine. And then she began to deal with people like parents, grandparents, and later on, we introduced a younger brother, Tian Tian, to play with her. So this would be a family of four. So about a year and a half ago, we let her go to a kindergarten, a community.

Right now, what you saw was the AI Town. When she needed to solve a much bigger task, she needed to find out who was responsible for a streamed cat, right? Not a scratched lady. So this is the abstract task she comes back and forth, and dealing with different types of people.

It was later confirmed that he had formally submitted his resignation to Alibaba on the afternoon of March 3. The news was subsequently shared within a small circle inside the Qwen team, prompting strong emotional reactions among team members.

The departure came very suddenly and no official explanation was provided, leaving outside observers temporarily uncertain about his next move or even whether he had fully left Alibaba.

At the same time, several other key members of the Qwen team also departed:

• Yu Bowen, head of post-training, left on the same day.

• Huibin, head of Qwen Code, had already joined Meta in January 2026.

These developments have been widely interpreted as an important signal of a wave of senior-level departures within the Qwen team.

An exclusive report by LatePost has revealed some behind-the-scenes details of the incident.

In the early hours of Wednesday, March 4, Lin Junyang, the technical lead of Alibaba’s Qwen team, suddenly posted a message on social media: “me stepping down. bye my beloved qwen.”

It was later confirmed that Lin had formally submitted his resignation to Alibaba on the afternoon of March 3. Shortly afterward, the Qwen team shared the news internally within a small circle. According to people familiar with the matter, some members of the Qwen team were deeply emotional upon hearing that he would be leaving—one colleague was even said to have “broken down in tears.”

As of 4:00 a.m. Beijing time, Lin Junyang’s social media post had received more than 5,000 likes and over 700 comments, most of them expressing appreciation for the contributions of the Qwen team and the open-source large model community.

On the same day, Yu Bowen, who had been responsible for post-training at Qwen, also formally departed. His role will be taken over by Zhou Hao, a former Senior Staff Researcher at DeepMind who joined Alibaba’s Tongyi Lab earlier this year. Zhou Hao reports to Alibaba Cloud CTO and Tongyi Lab head Zhou Jingren.

We have also learned that Huibin, the head of Qwen Code, had already left Alibaba in January 2026 to join Meta. Lin Junyang subsequently took over responsibility for Qwen Code, and as recently as last week he was still sharing recruitment posts on social media related to the Qwen Coding Agent.

Several people close to the matter told us that Lin’s departure came as a surprise. “There’s a sense of regret,” one said. “It’s bittersweet. He really loved Qwen.”

Lin’s decision to leave likely has something to do with the organizational restructuring currently underway within the Qwen team.

The Qwen (Tongyi Qwen) team directly overseen by Lin Junyang sits within Tongyi Lab, which is led by Alibaba Cloud CTO Zhou Jingren. Recently, Tongyi Lab has planned to split the Qwen team apart—moving away from a vertically integrated structure covering different training pipelines and modalities toward a horizontally structured system with separate teams for pre-training, post-training, text, multimodal, and other functions. These teams will remain under Tongyi Lab, but Lin’s management scope has been reduced.

This move to break up and redistribute the model teams also runs counter to Lin Junyang’s own view of where the technology is heading. Over the past year, Lin has repeatedly argued that pre-training, post-training, infrastructure, and training teams should be more tightly integrated and communicate more closely. LatePost previously reported that the Qwen model team began building its own infrastructure team from mid-last year; previously this work had largely been handled by Alibaba Cloud’s AI platform PAI, which supports infrastructure needs for multiple teams within Tongyi Lab.

Over the past one to two years, several major Chinese tech companies have carried out multiple rounds of adjustments to the organization of their model teams, with Alibaba making comparatively fewer changes. In comparison with other Chinese companies’ AI structures: ByteDance’s Seed division internally runs a “horse-race” system where different teams work on the same direction and modality; the Doubao main model series is organized around the workflow stages of pre-training and post-training; and Tencent, following adjustments last year, has become more integrated, with both model training and infrastructure teams under the management of Yao Shunyu.

Given Qwen’s reputation and achievements in recent years, Lin Junyang is unlikely to lack opportunities. Several investors and major companies had already been in contact with him, some hoping he would start a company of his own, while others had extended job offers.

Prior to this change, the Qwen team had already been facing subtle internal tensions within Alibaba.

On the one hand, Qwen enjoys strong support in the global open-source community. Its wide range of model sizes has made it popular among smaller startups, and a number of well-known companies—such as Cursor—fine-tune and post-train their products on top of Qwen models. Qwen’s open-source multimodal models have also become the base models of choice for many Chinese embodied-AI companies.

At the same time, Qwen and Lin Junyang have continued expanding the boundaries of the team’s capabilities, creating overlaps with other parallel teams inside Tongyi Lab. For example, Qwen has been developing VLA embodied models, while another team in Tongyi Lab led by Xu Zhuhong is working on similar efforts. Qwen has also been building text-to-image models (Qwen-image) and speech models, areas that overlap with Tongyi Lab’s Tongyi Wanxiang (focused on multimodal generation) and Bailin (focused on speech models). As Qwen also began building its own infrastructure teams, it gradually started to resemble a fully fledged “full-stack AI lab.”

On the other hand, Alibaba internally has been continuously evaluating Qwen’s results and value.

Some questions have been raised about the commercialization efficiency of open-source models: although Qwen enjoys strong reputation, open sourcing may affect Alibaba’s ability to generate direct revenue through selling model APIs.

There have also been internal assessments of specific Qwen outputs. We understand that some Alibaba executives were not fully satisfied with Qwen-3.5, which was unveiled on Lunar New Year’s Eve, describing it as a “half-finished product.”

From Alibaba’s broader perspective, technological influence and contributions to the open-source community are not ends in themselves, but rather means to achieve strategic and commercial goals such as AI cloud and a “super AI app.” In the AI cloud market, Alibaba Cloud is facing aggressive competition from ByteDance’s Volcano Engine, while ByteDance has adopted a closed-source model strategy. On the super-app front, during the recently concluded Lunar New Year subsidy battle, the Qwen app did not significantly narrow the gap with Doubao.

The commercial objectives and technological goals were not fully aligned. The tension between top-down strategic planning and division of responsibilities on the one hand, and the independent exploration of smaller internal teams on the other, reflects a deeper issue between the Qwen team and the broader environment within Alibaba.

Notably, Lin Junyang, Yu Bowen, and Huibin—who have all recently left the company—started their careers at Alibaba as fresh graduates and were trained within the company. Lin Junyang joined Alibaba DAMO Academy in 2019 after earning a master’s degree in Linguistics and Applied Linguistics from Peking University. Yu Bowen joined DAMO Academy in 2022 after completing his PhD at the Institute of Information Engineering of the Chinese Academy of Sciences and was one of Alibaba’s “Alibaba Stars” that year. Huibin, born in 1999, formally joined Alibaba DAMO Academy in 2022 after obtaining a master’s degree from Tianjin University. All three were involved in the early training of the Qwen model.

Lin Junyang has an interdisciplinary background spanning computational linguistics and AI. Unlike many high-profile AI leaders, he is not an overseas-returned PhD but a technologist who rose within China’s domestic research environment while building international influence. In 2025, Lin—born in 1993 and then 32 years old—became Alibaba’s youngest P10.

Zhou Hao, who recently joined Alibaba, received his PhD from the University of Wisconsin–Madison in 2019. According to his LinkedIn profile, he was a key contributor to projects including Gemini 3.0, Al Mode, DeepResearch, and Gemini 1.0, and led multi-step reinforcement learning work for Gemini 3.0.

Lin Junyang’s departure has sparked extensive discussion within the AI community. Many of his colleagues and practitioners in the field expressed regret and appreciation for his work on social media.

Based on accounts from several people close to Lin, his management style emphasized supporting team members and encouraging both self-motivation and cohesion. He believed an ideal manager should be a “reasonable person” who operates with logic. For leaders of smaller teams, he often said the most important responsibility is to recruit people better than themselves—otherwise it is “a failure.” Behind this philosophy, he believed leaders must keep their ego small, not assume they are invincible or capable of doing everything themselves.

A member of Tongyi Lab once told us that before the AI boom of 2023, the Qwen team—already developing large models at the time—had grown in a relatively unnoticed corner of the organization. With fewer interruptions or competing pressures, the team was able to focus its energy on iterating and improving the model itself.

But once AI became an all-or-nothing “total war” among major tech companies, the core model research teams at each company began facing far more organizational changes. Such changes are often triggered when R&D efforts encounter clear setbacks, yet Alibaba’s latest adjustment came at a time when both external evaluations and internal morale were relatively strong.

Lin Junyang’s decision to resign also came as a surprise to Alibaba. From the company’s perspective, however, the needs of the organization ultimately take precedence over the preferences of any individual.

Later today, 36Kr revealed details about an emergency all-hands meeting Alibaba held following Lin’s resignation.

“I should have known about this earlier.”

At around 13:00 Beijing time on March 4, Tongyi Lab convened an emergency All Hands meeting. Alibaba Group CEO Eddie Wu spoke candidly to employees of the Qwen team.

Twelve hours earlier (00:11 a.m. Beijing time on March 4), Lin Junyang, the technical lead of Alibaba’s Qwen large model, had suddenly announced his resignation on X. Lin is widely regarded as a central driving force behind Alibaba’s open-source AI models and one of the youngest P10 technical leaders at the company. As the news spread across the industry, some members of the Qwen team themselves struggled to accept the abrupt departure of such a pivotal figure.

“Given that we achieved today’s results with far fewer resources than our competitors, Junyang’s leadership was one of the key factors,” more than one Qwen team member told Intelligent Emergence.

During the meeting, several Qwen members—represented by Liu Dayiheng, the Qwen RL (reinforcement learning) lead—raised a series of questions to Alibaba’s senior leadership. These concerns touched on topics such as the planned team restructuring, the arrival of new member Zhou Hao, the direction of model development, and the allocation of resources.

Participants in the meeting included several Alibaba executives, members of the Qwen team, and staff from other teams within Tongyi Lab. Addressing issues related to organizational changes and strategic direction, Alibaba CEO Eddie Wu, Alibaba Chief Talent Officer Jiang Fang, and Alibaba Cloud CTO Zhou Jingren responded to multiple questions.

According to Alibaba’s leadership, the core characterization of the recent changes is that Qwen is not being downsized. Instead, they described the move as an expansion of the team, unrelated to any political infighting, and one that would require additional resources.

“We are growing rapidly. This adjustment is meant to bring in more talent and provide more resources,” said Alibaba Chief Talent Officer Jiang Fang. She also acknowledged shortcomings in communication: “We didn’t communicate this organizational change well enough. Bringing in new people will inevitably change the formation of the team, and expansion always involves adjustments. We may not have handled this properly.”

There had been speculation that Zhou Hao would directly lead Lin Junyang and his related teams. However, according to Intelligent Emergence, Zhou Hao’s exact role and reporting lines are still under discussion.

During the meeting, Alibaba executives repeatedly emphasized that the Qwen foundational model is currently the most important priority for the entire group. The competition in large models, they stressed, is not just a matter for the Qwen team but for the whole Alibaba organization. Both foundational model development and underlying infrastructure will be coordinated at the group level—“we must surpass the competition.”

Alibaba Cloud CTO Zhou Jingren also addressed several pointed questions, including issues around hiring quotas and shortages of computing power. Some team members asked why external clients—such as startup companies building large models—appear able to purchase Alibaba Cloud computing resources smoothly, while internal teams struggle with limited computing capacity and restricted hiring quotas.

Zhou responded that the team is currently operating under tight resource constraints, noting that the differences between internal and external allocation stem from a variety of historical reasons. He added that broader planning is underway but did not elaborate further.

As for Lin Junyang’s next move, no definitive update was given during the meeting. However, around 2 p.m. that afternoon, Lin posted another message on his social media feed, saying: “Qwen brothers, just keep working as planned. Everything’s fine.” He did not clarify whether he would be returning.

Just days earlier, Alibaba had completed a new round of updates to its AI strategy, unifying the umbrella name and core brand for its AI efforts under Qwen. Organizationally, the company also introduced another round of internal restructuring.

According to Intelligent Emergence, the Qwen team previously maintained its own pre-training, post-training, and infrastructure teams. In terms of model modalities, the team also covered several directions, including language models, multimodal models, and code models.

In the past, training single-modality models had been the industry norm. However, as demand for visual understanding continued to grow, vision–language models (VLMs) emerged, making deep integration across modalities a major trend in the field.

A person familiar with the situation told Intelligent Emergence that starting in 2025, Lin Junyang had been exploring ways to bring together employees working on language, image, video, and code models within Qwen, with the goal of improving training efficiency. At one point, the Qwen team proposed merging with the Wanxiang team, but the plan did not materialize; instead, the Qwen team eventually developed its own Qwen-image model.

In the latest round of restructuring, however, Tongyi Lab planned to split the Qwen team along functional lines such as pre-training, post-training, vision understanding, and image generation, merging these units with corresponding teams within Tongyi Lab—such as Tongyi Wanxiang and Tongyi Bailin—to work together. But with insufficient communication around the changes, tensions ultimately surfaced.

One day before the incident took place, Jack Ma unexpectedly appeared at a school in Hangzhou. What was particularly notable was that this time nearly all of Alibaba’s core executives were present, including Joe Tsai, Eddie Wu, Shao Xiaofeng, and Jiang Fan from Alibaba, as well as Eric Jing and Han Xinyi from Ant Group.

Together with them, Jack Ma spent more than an hour speaking with the school’s principal and teachers, discussing the challenges and opportunities brought about by AI.

Below is the official statement released by the school, which reveals many additional details.

On the day before the start of the new school term, the core leadership teams of Alibaba and Ant Group visited Yungu School.

Joe Tsai, Eddie Wu, Shao Xiaofeng, and Jiang Fan from Alibaba, together with Eric Jing and Han Xinyi from Ant Group, joined Mr. Ma in a discussion with the school’s principal and teachers that lasted for more than an hour. They exchanged views on the challenges and opportunities brought by artificial intelligence.

Mr. Ma told the group that the era of AI has already arrived rapidly, and its impact on society may exceed expectations. None of us, he said, is fully prepared for it. But for teenagers, the opportunities for change are the greatest. The purpose of this visit to Yungu was therefore to share Alibaba’s increasingly clear insights into AI with the educators.

During the discussion, participants noted that AI is evolving on a weekly cycle, with its capabilities continuing to grow. This technological revolution will bring historic changes to productivity and many aspects of society. Material wealth in the future could increase dramatically. People may no longer need to work eight hours a day, but many of the professions familiar today may disappear. Mr. Ma explained that the reason they came together to Yungu was to emphasize that this transformation will come very quickly, and education must adapt rapidly to help children learn to coexist with AI and adjust to this profound shift.

Joe Tsai, Chairman of Alibaba Group, remarked that in the AI era, critical thinking will become increasingly important. Critical thinking is not simply about asking questions, but about asking the right questions. In the future, machines may be able to perform many tasks, but communication—both between humans and machines and among people themselves—could become one of the most important capabilities.

Alibaba CEO Eddie Wu said that the key differences between humans and machines in the future lie in three qualities: curiosity, empathy, and physical capability. Curiosity drives people to explore and act voluntarily, whereas machines are passive. Empathy reflects the ability to understand others. And when intellectual tasks are increasingly handled by AI, physical ability will become more important, meaning sports and physical education will gain greater significance.

Eric Jing, Chairman of Ant Group, said that AI should be used to handle repetitive and tedious work so that people have more time to develop their unique human qualities, including aesthetic sense, creativity, and imagination. At the same time, while making good use of AI, we should avoid letting it become a crutch we cannot discard, and must preserve the ability to think independently.

Jack Ma added that while AI brings tremendous disruption, it also creates enormous opportunities—especially the opportunity for education to return to its true essence. Time previously spent on rote memorization and endless test practice can be freed up and redirected toward nurturing creativity and imagination. Children will have more time to play, to learn music, painting, and sports, and through these experiences they can learn to share, to feel and experience the world, to listen, and to understand others.

Participants also discussed how Yungu School might adapt to these changes in the future. Mr. Ma noted that whether a school belongs to the AI era should not be judged by how many AI servers it has or how strong its AI technical capabilities are. AI possesses “chips,” but humans possess hearts. The greatest change AI will bring to education is that teachers can fully embrace their role as “engineers of the soul,” rather than mere transmitters of knowledge. The goal is not for children to compete with AI in computation or memory, but to remain curious, develop empathy and responsibility, and cultivate a sense of lived experience. Curiosity, imagination, creativity, judgment, and aesthetic sensibility are the true abilities that education must foster in children in the age of AI.

Bloomberg Reporter

In light of the situation in Iran and President Trump’s expected visit to China in the near future, how do you view the trajectory of China–U.S. relations? In particular, considering that more leaders’ meetings are scheduled later this year.

Lou Qinjian

Thank you for the question. China has consistently maintained that, as the world’s two major powers, China and the United States should respect each other, coexist peacefully, and pursue win-win cooperation. President Xi Jinping has pointed out that for the two countries to be partners and friends is both a lesson of history and a necessity of reality. The notion that China and the United States can “achieve mutual success and shared prosperity” is a tangible reality.

Head-of-state diplomacy plays an irreplaceable strategic guiding role in China–U.S. relations. Since last year, President Xi Jinping and President Trump have maintained regular communication, helping to steer the great ship of China–U.S. relations on a steady course and inject new momentum into it. Overall, bilateral relations have remained stable and improving, a development widely welcomed by both countries and by the international community. Facts have shown that when China and the United States cooperate, both benefit; when they confront each other, both suffer. As long as both sides faithfully implement the important consensus reached by the two heads of state, uphold equality, mutual respect, and reciprocity, expand the list of cooperation and narrow the list of problems, China–U.S. relations can move forward in a stable manner.

China stands ready to work with the United States to strengthen communication at all levels and through multiple channels, opening broader space for cooperation. At the same time, China has its own principles and bottom lines, and will as always firmly safeguard its sovereignty, security, and development interests. Last year, the National People’s Congress invited a delegation of U.S. House members to visit China, maintained dialogue with U.S. senators, and sent a youth delegation of NPC representatives to the United States. A common sentiment expressed by U.S. lawmakers was that meeting is better than not meeting, and dialogue is better than no dialogue. The National People’s Congress is willing to maintain exchanges with the U.S. Congress and contribute, from the perspective of legislative institutions, to the stable and healthy development of China–U.S. relations. We hope the U.S. Congress will view China and its development objectively and rationally, and do more to promote friendship between the two countries and their peoples. Thank you.

Czech Media Reporter

Last year, political leaders from several European countries and heads of EU institutions visited China. Some observers say Europe is “turning toward China,” while others argue that Europe faces competition and challenges from China in areas such as trade and economics. How do you view this?

Lou Qinjian

Thank you for your question. Last year marked the 50th anniversary of diplomatic relations between China and the European Union, and high-level exchanges between the two sides were frequent. President Xi Jinping met with European leaders many times, reaching important consensus on strengthening the China–EU partnership, deepening mutually beneficial cooperation, and jointly addressing global challenges. Practical cooperation between China and Europe has yielded fruitful results, and people-to-people exchanges have been vibrant. The facts show that Europe needs China, and China also needs Europe.

The essence of China–EU economic and trade relations is complementarity and mutual benefit. Cooperation with China helps Europe develop its economy and improve people’s well-being. For example, the China–Europe Trade and Logistics Cooperation Park in Hungary has created a platform for two-way interaction between Chinese and European companies, further strengthening Hungary’s position as a logistics hub in Europe. The Pelješac Bridge in Croatia, funded by EU funds and built by Chinese companies, has connected north and south, turning a natural barrier into a corridor and fulfilling a long-held aspiration of local residents.

Another example is China’s State Grid investment in Portugal’s national energy network company. By sharing China’s expertise in grid operation management and renewable energy integration, it has supported Portugal’s energy transition and achieved mutual learning and win-win cooperation. PetroChina and Shell jointly invested in Arrow Energy in Australia, combining their strengths to provide innovative, efficient, and environmentally friendly energy solutions for a third country, while also creating local employment and promoting economic growth.

Similarly, European companies investing in China also benefit from China’s high-quality development. For instance, BASF’s integrated base in Zhanjiang is the company’s largest single investment to date, while Sanofi’s insulin raw-material project in Beijing has become its largest insulin injection production base in the Asia-Pacific region. While meeting the needs of the Chinese market, these companies have also significantly enhanced their global competitiveness.

There are no fundamental conflicts of interest or geopolitical contradictions between China and Europe. China is willing to work with Europe to maintain the basic positioning of partnership, properly handle trade differences, expand cooperation agendas, and jointly respond to global challenges. Last year, China and the European Parliament simultaneously lifted restrictions on exchanges, restarting long-suspended institutional communication between legislative bodies. The National People’s Congress is willing to strengthen friendly exchanges with the European Parliament and national parliaments across Europe, deepen understanding, build consensus, expand cooperation, and contribute to the steady and long-term development of China–EU relations. Thank you.

Lianhe Zaobao Reporter (Singapore)

Over the past year, China’s high-level exchanges with neighboring countries have been frequent and regional cooperation has shown positive momentum. However, difficulties and setbacks in China–Japan relations have also drawn attention. In this context, how will China promote the development of relations with neighboring countries, including Japan?

Lou Qinjian

Thank you for your question. Neighboring countries are neighbors that cannot be moved away. How a major power interacts with its neighbors reflects its worldview, its concept of order, and its values. China has always placed its neighborhood at the top of its overall diplomatic agenda and has consistently been committed to promoting regional peace, stability, development, and prosperity. President Xi Jinping has visited almost all neighboring countries, and last year’s major diplomatic events and important visits also focused on the region. As you have seen, under the strategic guidance of head-of-state diplomacy, China and its neighbors have strengthened strategic mutual trust, deepened comprehensive cooperation, and expanded exchanges across various fields, injecting strong momentum into regional stability and development.

Safeguarding national sovereignty, security, and development interests is the starting point and ultimate goal of all foreign relations. The Taiwan question is China’s internal affair and the core of China’s core interests. The Chinese side firmly opposes erroneous remarks by Japanese leaders related to Taiwan. The Chinese people will never allow any external forces to interfere in China’s internal affairs and will resolutely safeguard national sovereignty, unity, and territorial integrity.

At present, maintaining stability, improving governance, and promoting development remain the mainstream aspirations in the region, and the trend toward regional integration is irreversible. China will fully implement the outcomes of the Central Conference on Work Relating to Neighboring Countries, focus on building a community with a shared future with neighboring countries, adhere to the principles of amity, sincerity, mutual benefit, and inclusiveness, and work with neighboring countries to build a peaceful, secure, prosperous, beautiful, and friendly home. The National People’s Congress will continue strengthening bilateral and multilateral exchanges with legislative bodies in neighboring countries, helping China’s modernization and the development of neighboring countries advance together and complement each other.

Reporter, Anadolu Agency (Türkiye)

The United States and Israel recently launched military actions against Iran, resulting in the deaths of its leaders. Not long ago, the world witnessed the president of a sovereign state being arrested in Venezuela. The United States has also long expressed interest in annexing Greenland, which is Danish territory in the Arctic. As unilateralism and the use of force become more prominent in international relations, the international community is increasingly concerned about the collapse of the global governance system and the weakening role of the United Nations. What is your view?

Lou Qinjian

Thank you for the question. China is paying close attention to the situation in Iran. Iran’s national sovereignty, security, and territorial integrity should be respected. China calls for an immediate halt to military actions, avoiding further escalation of tensions, restoring dialogue and negotiations, and safeguarding peace and stability in the Middle East. China stands ready to continue playing the role of a responsible major country.

China has always believed that mutual respect and equality among countries, regardless of size, are requirements of historical progress and the primary principle of the UN Charter. No country has the right to control international affairs, determine the fate of other nations, or monopolize development advantages, still less to act arbitrarily on the world stage.

Over the past year, the evolution of the international landscape has reached an important turning point. Economic globalization has suffered severe shocks, while regional wars and cross-border conflicts have occurred frequently. In the face of a turbulent world, peaceful development, cooperation, and mutual benefit remain the only right path. President Xi Jinping has solemnly put forward initiatives on global governance, emphasizing sovereign equality, respect for international law, the practice of multilateralism, a people-centered approach, and an action-oriented agenda. These proposals present China’s plan for building a fairer and more reasonable global governance system.

Last year marked the 80th anniversary of the victory in the World Anti-Fascist War and also the 80th anniversary of the founding of the United Nations. The devastation of the two world wars prompted the international community to reflect deeply, leading to the creation of the United Nations. More than 80 years of practice has shown that the role of the United Nations must be strengthened, not weakened, and its position must be upheld, not replaced. As a major country and a permanent member of the UN Security Council, China is willing to work with all countries to firmly uphold the authority and status of the United Nations, safeguard the purposes and principles of the UN Charter, practice genuine multilateralism, improve global governance, and contribute more to building a community with a shared future for mankind. Thank you.

He reviewed SHEIN’s development since establishing operations in Guangzhou in 2014 and expressed gratitude to the Guangdong Provincial Party Committee, the provincial government, and various departments at all levels for their long-standing support.

He said that leveraging Guangdong’s comprehensive apparel industry supply chain, mature cross-border e-commerce ecosystem, and highly efficient international logistics system, SHEIN built its “small-batch, rapid replenishment” model, achieving an operational cycle of just two to three weeks from design to delivery. The company now serves more than 160 countries and regions worldwide, with cumulative platform export value exceeding RMB 100 billion. Guangdong’s first-class business environment and its “non-interference unless necessary, full support when needed” service philosophy have provided strong institutional backing for enterprise growth. As a result, Guangdong has become SHEIN’s supply chain headquarters and its most important development base, supporting nearly 10,000 partner suppliers and generating hundreds of thousands of jobs.

Chris Xu also summarized three key lessons from the company’s growth:

First, industrial ecosystems and the business environment are the foundation for enterprise expansion.

Second, digitalization serves as the bridge integrating manufacturing and services—driving production with data and optimizing manufacturing through real-time market feedback to build a competitive advantage of “speed plus precision.”

Third, as a supply chain leader, SHEIN proactively assumes social responsibility by empowering Guangdong manufacturing to expand globally with higher quality.

According to Chris Xu, SHEIN will continue increasing investment in Guangdong, deepening integration between cross-border e-commerce and industrial clusters, opening digital tools to partners, strengthening talent training and smart supply chain development, and investing over RMB 10 billion to build a world-class fashion industry cluster—helping “Guangdong products reach the world” while achieving mutual growth with the province in the course of high-quality development.

Below is the video recording and full text of Chris Xu’s speech:

Distinguished Secretary Kunming, Governor Fanli, respected leaders, fellow entrepreneurs, Happy New Year.

Standing here today as a representative of enterprises, I feel deeply grateful and inspired. Every step of SHEIN’s growth has been nurtured by this vibrant land of Guangdong. Since establishing ourselves in Guangzhou in 2014, we have maintained rapid growth, achieving cumulative platform export value exceeding RMB 100 billion. Our product offerings now cover apparel, footwear, beauty, and home goods, serving customers in over 160 countries and regions and becoming a globally recognized fashion retailer. On behalf of SHEIN Group, I would like to extend my sincerest appreciation to all government leaders and partners who have supported us.

Allow me to share three reflections from SHEIN’s journey:

First, Guangdong’s comprehensive industrial ecosystem and first-class business environment have enabled SHEIN’s rapid growth.

Our connection with Guangdong began with deep recognition of its industrial ecosystem. From apparel factories in Panyu to logistics hubs in Baiyun, and the cross-border ecosystem spanning the province, these elements form the backbone of our small-batch, rapid replenishment model. From design to final delivery into customers’ hands, we can compress the entire cycle to two to three weeks—reflecting the ultimate efficiency of Guangdong’s apparel supply chain and the high performance of its international logistics system.

What has touched us even more is Guangdong’s world-class business environment. SHEIN’s achievements would not have been possible without the care and guidance of the provincial Party Committee and government, the strong support of the Guangzhou municipal leadership, and the precise assistance from departments such as the provincial and municipal commerce bureaus, customs, and taxation authorities. When we first arrived in Guangzhou, leaders at provincial, municipal, and district levels proactively helped coordinate supply chain resources and implement supportive policies. This “non-interference unless necessary, full support when needed” approach gave us the confidence to base our supply chain headquarters in Guangzhou. Today, SHEIN works with nearly 10,000 suppliers in Guangdong, supporting over 600,000 jobs across the province. Guangdong has truly become our fertile ground for development.

Second, deep integration between manufacturing and services has enabled SHEIN’s global business expansion.

Leveraging the new cross-border e-commerce model and digitalization as a bridge, SHEIN embeds user insights, supply chain responsiveness, and global service networks throughout the manufacturing value chain, exploring new pathways for integrating manufacturing and services.

On the manufacturing side, demand has become the core engine driving production. Under the small-batch, rapid replenishment model, we use technology and digital analytics to capture global fashion trends, forecast industry movements, and translate fragmented market demand into frontline production instructions.

On the service side, we deeply integrate cross-border logistics and localized operations with manufacturing, using real-time sales data to guide factory production allocation, dynamically adjusting logistics delivery and task distribution across the chain. Customer feedback can directly trigger replenishment or product optimization, driving manufacturing upgrades in reverse.

It is precisely this integration—manufacturing embedded in services, and services driving manufacturing—that has enabled SHEIN to build a dual moat of speed and precision, becoming a leading global fashion enterprise.

Third, as a supply chain leader, SHEIN is committed to supporting the high-quality global expansion of Guangdong manufacturing.

We regard serving the high-quality development of Guangdong manufacturing as our core mission. SHEIN’s digital factory toolkits are fully open across the supply chain, helping factories achieve information transparency and refined management. We have built smart supply chain networks in Guangzhou, Foshan, Zhaoqing, and Jiangmen, stimulating growth in warehousing and logistics industries. We have also established an apparel manufacturing innovation center as a shared training base integrating industry, education, and evaluation. In 2025 alone, we conducted nearly 600 training sessions, benefiting 37,000 participants from supplier companies, strengthening the talent foundation for high-quality development.

Looking forward, SHEIN will further strengthen its responsibility as a supply chain leader. Since launching the “High-Quality Industrial Belt Global Expansion Plan,” Guangdong has become a model province, achieving full coverage of all 21 prefecture-level cities in 2024. Over the next three years, we will actively participate in “cross-border e-commerce + industrial belt” pilot programs, enabling more SMEs to benefit from cross-border trade and helping Guangdong products reach global markets steadily and sustainably. We will continue investing in supply chain empowerment and talent development, advancing digital transformation in traditional manufacturing and supplying more interdisciplinary talent to the Greater Bay Area. We will remain rooted in Guangdong, investing over RMB 10 billion to build SHEIN’s smart supply chain system and jointly create a world-class fashion industry cluster.

Distinguished leaders and colleagues, Guangdong is SHEIN’s root and the starting point of our journey. We firmly believe that amid Guangdong’s high-quality development wave, SHEIN will grow together with this vibrant land, making “Guangdong Intelligent Manufacturing” a new benchmark for the global fashion industry.

Thank you.

As Chris Xu noted, SHEIN’s rapid growth stems from its unique business model—flexible, on-demand production; small-batch, rapid replenishment; and direct-to-consumer sales.

Specifically, SHEIN typically launches small batches of 100–200 units to test customer response before deciding whether to scale production. This highly automated process ensures alignment between demand and supply while minimizing overproduction. SHEIN’s production waste rate remains in the single digits—far below industry averages. Behind this on-demand production model lies deep digital transformation of the supply chain.

SHEIN operates in close coordination with its suppliers. Through SHEIN’s proprietary supply chain technology solutions suite, suppliers gain visibility into production capacity, inventory levels, and demand signals, enabling more informed manufacturing decisions. SHEIN also invests in suppliers’ technological upgrades to ensure compatibility with its digital systems. Suppliers’ ability to assess demand signals and optimize production processes is central to SHEIN’s agile, efficient, small-batch approach. Meanwhile, SHEIN’s direct-to-consumer sales strategy enables rapid payment cycles, far shorter than industry norms. The company maintains long-term relationships with more than 5,000 third-party contract manufacturers.

With high levels of coordination across the supply chain, SHEIN’s design and production speed outpaces other fast fashion brands. Core suppliers are geographically concentrated, primarily in Guangzhou. Suppliers must complete design and production within 10 days—significantly faster than Zara’s well-known three-week turnaround time.

Originally incorporated in Nanjing, SHEIN later changed its operating entity to Guangzhou SHEIN International Import & Export Co., Ltd., controlled by Hong Kong–based Zoetop Business Co. At the end of 2021, ultimate control shifted to Singapore-based Roadget Business Pte. Ltd., making Singapore the company’s headquarters. The previously financed Nanjing Lingtian Information Technology Co., Ltd. has since been deregistered. According to Public Eye, both Zoetop Business and Roadget Business are owned by Beauty of Fashion Investment, a British Virgin Islands entity.

Although SHEIN has formally distanced itself from China in corporate structure, it continues to source nearly all products from China. These Chinese suppliers underpin the company’s small-batch, rapid replenishment model. In the short term, SHEIN cannot easily relocate its supply chain and therefore remains closely tied to China.

Among them, 30 entities were added to the Export Control List (all defense-related companies; Chinese companies are prohibited from conducting transactions with them), while 20 entities were added to the “Watch List” (midstream and supporting enterprises in the defense supply chain. It is similar to the U.S. Unverified List (UVL): Chinese companies may transact with them, but the relevant security review and risk assessment requirements are extremely stringent).

The two announcements:

MOFCOM Announcement No. 11 of 2026

Decision to Add 20 Japanese Entities to the Export Control List

Issuing Authority: Bureau of Security and Control
Document No.: MOFCOM Announcement No. 11 of 2026
Date of Issuance: February 24, 2026

Pursuant to the Export Control Law of the People’s Republic of China and the Regulations of the People’s Republic of China on Export Control of Dual-Use Items, and in order to safeguard national security and interests and fulfill international non-proliferation obligations, it has been decided to add 20 Japanese entities, including Mitsubishi Heavy Industries Shipbuilding Co., Ltd., that are involved in enhancing Japan’s military capabilities, to the Export Control List (see Annex). The following measures are hereby imposed:

1. Export operators are prohibited from exporting dual-use items to the above-mentioned 20 entities. Overseas organizations and individuals are prohibited from transferring or providing dual-use items originating from the People’s Republic of China to the above-mentioned 20 entities. Ongoing related activities shall be immediately suspended.

2. Where export is genuinely necessary under special circumstances, export operators shall apply to the Ministry of Commerce for approval.

This Announcement shall take effect as of the date of its publication.

Annex: Export Control List (February 24, 2026)

Ministry of Commerce
February 24, 2026

Annex

Export Control List

(February 24, 2026)

1. Mitsubishi Heavy Industries Shipbuilding Co., Ltd.

2. Mitsubishi Heavy Industries Aero Engines, Ltd.

3. Mitsubishi Heavy Industries Marine Machinery & Equipment Co., Ltd.

4. Mitsubishi Heavy Industries Engine & Turbocharger, Ltd.

5. Mitsubishi Heavy Industries Maritime Systems, Ltd.

6. Kawasaki Heavy Industries Aerospace Systems Company

7. KAWAJU Gifu Engineering Co., Ltd.

8. Fujitsu Defense & National Security, Ltd.

9. IHI Power Systems Co., Ltd.

10. IHI Master Metal Co., Ltd.

11. IHI Jet Service Co., Ltd.

12. IHI Aerospace Co., Ltd.

13. IHI Aero Manufacturing Co., Ltd.

14. IHI Aerospace Engineering Co., Ltd.

15. NEC Network and Sensor Systems, Ltd.

16. NEC Aerospace Systems, Ltd.

17. Japan Marine United Corporation

18. JMU Defense Systems Co., Ltd.

19. National Defense Academy of Japan

20. Japan Aerospace Exploration Agency

    ---

MOFCOM Announcement No. 12 of 2026

Decision to Add 20 Japanese Entities to the Watch List

Issuing Authority: Bureau of Security and Control
Document No.: MOFCOM Announcement No. 12 of 2026
Date of Issuance: February 24, 2026

Pursuant to the Export Control Law of the People’s Republic of China and the Regulations of the People’s Republic of China on Export Control of Dual-Use Items, it has been decided to add 20 Japanese entities, including SUBARU Corporation, for which the ultimate end users and end uses of dual-use items cannot be verified, to the Watch List (see Annex).

Where export operators export dual-use items to the above-mentioned entities, they shall not apply for a general license nor obtain export documentation through information filing and registration procedures. When applying for an individual license, they shall submit:

• A risk assessment report concerning the entity listed on the Watch List; and

• A written undertaking that the dual-use items will not be used for any purposes that contribute to enhancing Japan’s military capabilities.

The time limit for license review shall not be subject to the time limits specified in Article 17, Paragraph 1 of the Regulations of the People’s Republic of China on Export Control of Dual-Use Items.

The Ministry of Commerce shall implement stricter end-user and end-use reviews for exports of dual-use items to entities on the Watch List. Exports involving Japanese military end users, military end uses, or any other end users or end uses that contribute to enhancing Japan’s military capabilities shall not be approved.

Entities included on the Watch List may, pursuant to Article 26 of the Regulations of the People’s Republic of China on Export Control of Dual-Use Items, apply for removal from the Watch List after fulfilling their obligation to cooperate with verification. Upon verification, the Ministry of Commerce may remove them from the Watch List.

This Announcement shall take effect as of the date of its publication.

Annex: Watch List (February 24, 2026)

Ministry of Commerce
February 24, 2026

Annex

Watch List

(February 24, 2026)

1. SUBARU Corporation

2. FUJI Aerospace Technology Co., Ltd.

3. ENEOS Corporation

4. Yusoki Co., Ltd.

5. ITOCHU Aviation Co., Ltd.

6. Leda Group Holdings Co., Ltd.

7. Institute of Science Tokyo

8. Mitsubishi Materials Corporation

9. ASPP Co., Ltd.

10. Yashima Denki Co., Ltd.

11. Sumitomo Heavy Industries, Ltd.

12. TDK Corporation

13. Mitsui Bussan Aerospace Co., Ltd.

14. Hino Motors, Ltd.

15. Tokin Corporation

16. Nissin Electric Co., Ltd.

17. Sun Tectro Co., Ltd.

18. Nitto Denko Corporation

19. NOF Corporation

20. Nacalai Tesque, Inc.

Why are these companies targeted? A Guancha piece explained a bit.

In December last year, the Stockholm International Peace Research Institute (SIPRI) released its 2024 ranking of the world’s top 100 arms-producing and military services companies. Several Japanese firms, including Mitsubishi Heavy Industries and Kawasaki Heavy Industries, recorded substantial increases in defense-related revenues. This reflects what appears to be a gradual but potentially risky loosening of Japan’s arms export policy.

Mitsubishi Heavy Industries (MHI), Kawasaki Heavy Industries (KHI), and IHI (formerly Ishikawajima-Harima Heavy Industries) are widely regarded as the three pillars of Japan’s postwar heavy industrial complex. Among the entities placed on the control list, the Mitsubishi group occupies the most prominent position, with five affiliated companies included.

Mitsubishi Heavy Industries is Japan’s largest comprehensive defense contractor and is also advancing next-generation nuclear energy technologies. The listed subsidiaries cover key segments of the defense industrial chain, including shipbuilding, aero engines, naval propulsion systems, and maritime systems integration.

Mitsubishi Shipbuilding traces its origins to the Mitsubishi Nagasaki Shipyard, established in 1884. During World War II, it built major warships such as the battleship Musashi and served as a core supplier to the Imperial Japanese Navy. Today, Mitsubishi Shipbuilding continues to participate in the construction of submarines for the Japan Maritime Self-Defense Force, Atago-class Aegis destroyers, and Mogami-class frigates.

Mitsubishi Heavy Industries Aero Engines is a core player in Japan’s aviation power sector and is responsible for maintaining the propulsion systems of Japan Air Self-Defense Force fighter aircraft. Mitsubishi Heavy Industries Marine Machinery provides ship machinery and equipment, including large-scale engines and propulsion systems. Mitsubishi Heavy Industries Engine & Turbocharger produces high-performance engines and turbochargers, some of which are used in Ground Self-Defense Force missile systems. Mitsubishi Heavy Industries Maritime Systems develops combat management systems for naval vessels, airborne radar systems for anti-submarine patrol aircraft, and submarine sonar systems.

Two Kawasaki Heavy Industries subsidiaries were also listed. Its Aerospace Systems Company has long served the Japan Air Self-Defense Force. Major platforms such as the C-2 transport aircraft, the P-1 maritime patrol aircraft, and various helicopters have been led or co-produced by Kawasaki. KAWAJU Gifu Engineering undertakes extensive aerospace design, manufacturing, testing, and maintenance work, supporting both active-duty equipment and next-generation aviation platforms. Structurally, the Kawasaki group forms a key material foundation of Japan’s air defense capabilities.

IHI is another major Japanese defense contractor and a core supplier of engines, rocket systems, and advanced materials. Six IHI subsidiaries were included on the control list, covering businesses such as large gas turbine power systems, high-performance metallic materials for aero engines, and maintenance and support services for military jet engines. IHI Aerospace also participates in rocket engine development and spacecraft structural projects.

Japan Marine United (JMU) is one of Japan’s principal civilian and military shipbuilders. The Izumo-class helicopter carrier, currently the largest operational surface combatant in the Japan Maritime Self-Defense Force, was built by JMU.

The Chinese side emphasized in its statement that the purpose of this round of measures is to curb Japan’s “remilitarization” and its alleged pursuit of nuclear capabilities, and that the actions are entirely justified, reasonable, and lawful. China’s decision to place these entities on the lists in accordance with the law targets only a small number of Japanese entities. The relevant measures apply solely to dual-use items and do not affect normal China–Japan economic and trade relations. Japanese entities that operate in good faith and in compliance with the law have no cause for concern.

NHK Reporter: China’s Ministry of Commerce announced today that it has added 20 Japanese entities to its export control list and another 20 to its watch list. What is the Foreign Ministry’s comment on these measures? Why introduce them at this time?

Mao Ning: For specific questions, I would refer you to the competent Chinese authorities. In order to safeguard national security and interests and to fulfill international non-proliferation obligations, China has taken measures in accordance with the law and relevant regulations. The objective is to curb Japan’s “remilitarization” and its alleged pursuit of nuclear capabilities. These actions are entirely justified, reasonable, and lawful.

Japan’s Foreign Ministry said Masaaki Kanai, head of the Asian and Oceanian Affairs Bureau, spoke with China and “strongly protested and demanded the withdrawal of these measures.” Kanai called them “absolutely unacceptable and deeply regrettable,” the ministry said, adding that they “deviate significantly from international practice.”

Demarche by Director-General KANAI of the Asian and Oceanian Affairs Bureau of the Ministry of Foreign Affairs to Chinese Deputy Chief of Mission SHI

February 24, 2026

On February 24, following the announcement by the Ministry of Commerce of the People’s Republic of China, including the prohibition on exports of dual-use items to certain Japanese companies, on the same day, Director-General KANAI of the Asian and Oceanian Affairs Bureau, strongly protested and demanded the withdrawal of these measures to Chinese Deputy Chief of Mission SHI Yong, since these measures are absolutely unacceptable and deeply regrettable, in the same way as the measures announced on January 6 this year, stating that Japan has been protesting and demanding the withdrawal of the measures announced on January 6, which only target Japan, deviate significantly from international practice, and are absolutely unacceptable.

In a piece by Yu Yuan Tan Tian, the CCTV-affiliated outlets cited its discussion with Ji Wenhua, Professor of China’s UIBE and former official of MOFCOM’s Treaty and Law Department.

How should we assess the Chinese Ministry of Commerce’s latest measures? Tan Zhu discussed the issue with Ji Wenhua, a professor at the University of International Business and Economics. Three notable features stand out.

First, this marks the first activation of the “Watch List” mechanism, opening a new path for more targeted governance in China’s export control regime.

Article 26 of the Regulations on Export Control of Dual-Use Items, which came into effect on December 1, 2024, stipulates that where the ultimate end user or end use of dual-use items cannot be verified, the competent commerce authority of the State Council may place the relevant importer or end user on a Watch List.

The inclusion of 20 Japanese entities—such as SUBARU Corporation, the Institute of Science Tokyo, FUJI Aerospace Technology, and Mitsubishi Materials—represents the first practical application of this provision and carries important institutional significance.

The logic behind the Watch List is clear: by strengthening scrutiny and risk assessment, it enables precise regulatory control over entities whose end use is unclear and may pose potential military risks.

It is not an outright ban, but rather a regime of heightened review.

Moreover, this is not a permanent designation. The Ministry of Commerce has explicitly stated that listed entities may apply for removal if they genuinely fulfill their obligation to cooperate with verification. Once confirmed, they may be removed from the list in accordance with the law.

According to Ji Wenhua, the availability of reasonable and workable remedies reflects the rule-of-law spirit embedded in China’s control measures.

Second, the strict prohibitive measures imposed on entities placed on the Control List underscore China’s firm stance against Japan’s “remilitarization” and alleged pursuit of nuclear capabilities.

For the 20 entities included on the Control List—such as Mitsubishi Heavy Industries Shipbuilding—which are deemed to be involved in enhancing Japan’s military capabilities, the measures operate on two levels:

First, Chinese exporters are prohibited from exporting dual-use items to these entities.

Second, overseas organizations and individuals are prohibited from transferring or providing dual-use items of Chinese origin to these entities. Any ongoing related activities must cease immediately.

These measures are grounded in Articles 29 and 49 of the Regulations on Export Control of Dual-Use Items and are fully supported by law. They enable effective multi-layered control over both “direct exports” and “indirect transfers” of dual-use items, and they also demonstrate the extraterritorial application of China’s export control legislation.

Third, these measures represent an upgraded extension of Announcement No. 1.

On January 6, 2026, the Ministry of Commerce issued the Announcement on Strengthening Export Control of Dual-Use Items to Japan (Announcement No. 1 of 2026).

That announcement was issued in response to public statements by Japanese leaders regarding Taiwan, which implied the possibility of military involvement in the Taiwan Strait, constituted interference in China’s internal affairs, and seriously violated the One-China principle—acts deemed extremely egregious in both nature and impact.

Announcement No. 1 stated that, in order to safeguard national security and interests and fulfill international non-proliferation obligations, China decided to prohibit the export of all dual-use items to Japanese military end users, for military end uses, or for any other end use that would contribute to enhancing Japan’s military capabilities. Organizations and individuals in any country or region that violate these provisions by transferring or providing Chinese-origin dual-use items to Japanese entities will be held legally accountable.

The current listing of Japanese entities therefore constitutes an upgraded measure building upon Announcement No. 1, extending and deepening its scope in order to achieve multi-layered regulation, targeted prevention, and effective countermeasures against military-related Japanese entities.

This is possible because China has now largely established a modern export control legal framework, centered on the Export Control Law, supported by administrative regulations such as the Regulations on Export Control of Dual-Use Items, and coordinated across multiple ministerial rules.

Its defining feature is a clear hierarchical structure and coordinated design.

The Ministry of Commerce’s simultaneous use of a “dual-list” mechanism against Japanese entities represents the latest practice in balancing development and security within China’s export control system.

It builds a solid legal foundation for regulatory action, enhances governance efficiency through precise and tiered enforcement tools, diversifies China’s export control instruments through the first use of the Watch List mechanism, and fulfills international obligations in line with fundamental principles of international law.

At a time when globalization faces headwinds and regional security risks are intensifying, China’s policy toolbox is also expanding rapidly. In responding to actions that infringe upon China’s legitimate interests, China stands ready to take countermeasures at any time.

In the performance WuBOT（武 BOT）, robots developed by Unitree Robotics appeared alongside students from Henan Tagou Martial Arts School. With agile movements and seamless human–robot sparring, the show won thunderous applause. The robots demonstrated parkour, consecutive single-leg backflips, high-difficulty jumps, rapid formation changes, and even engaged in stick fighting and “drunken boxing” routines with the young performers.

Although it mainly showcased significant advances in hardware and balance control, it once again made the world take notice of Unitree’s technological progress and China’s growing strength in robotics.

Enjoy the whole show below:

The Spring Festival Gala was a breakout moment, and capital suddenly realized that a robot’s primary business might actually be performing. Trade shows, stages, competitions, celebrity concerts — these seem far more effective at attracting the next round of funding and generating buzz than putting robots to work in factories or real-world service roles.

Meanwhile, the large-scale deployment and commercialization of humanoid robots still face several formidable hurdles: reliability, production capacity, stability, and cost. As one industry observer put it, the Spring Festival Gala may be the perfect “comfort zone” for embodied intelligence. Factories, homes, retail spaces, workshops, and farmland all involve complex terrain and structure. Outdoor environments add uncontrollable variables such as lighting, humidity, temperature, and wind speed. Most importantly, robots must respond flexibly to unpredictable situations the way humans do — and technically, there are still many gaps to bridge.

Even a powerhouse like Tesla’s Optimus has faced credibility issues. In one widely circulated incident, a robot with no visible eye-mounted device suddenly raised its right hand and performed a precise, fluid “headset removal” motion. Market speculation suggested that a remote operator may have been controlling it, and the movement resembled the standard gesture of removing a VR headset. It was speculated that a device malfunction or overheating forced the operator to remove the headset in a rush without disconnecting, fueling public skepticism about whether embodied AI is truly intelligent — or merely a marionette. The episode also intensified doubts about whether some demonstrations are more about hype and fast fundraising than real capability.

Robots may struggle in real-world environments, but on stage they hold all the advantages. The Spring Festival Gala stage offers perfectly flat, high-friction flooring; stable indoor humidity and airflow; no exposure to weather; bright, preconfigured lighting; and even millisecond-level synchronization with background music. With pre-programmed routines — and possibly remote control — robots do not need to “think” or adapt. They simply execute a predetermined trajectory.

Last August in Beijing, I attended the World Robot Conference. And just last weekend, Beijing hosted the 2025 World Humanoid Robot Games (WHRG). My big takeaway was: hardware has clearly improved—motors, control systems, battery management are all much better. From the exhibition, you can see robots have made some real progress in autonomy. Some of the top ones can now carry out tasks completely on their own — spotting specific objects, making decisions, and manipulating them with high precision. Some robots really stood out in motor control and battery management.

But the “brains” haven’t kept up, which is the major shortcoming. Large models for embodied intelligence are still very inadequate, which makes robots look more like fancy toys than real labor replacements. Unitree’s CEO Wang Xingxing raised this issue in a keynote speech.

I also read reports from Goldman Sachs about the conference. I generally agree with their take: there’s still a lot of hype here, nowhere near mature enough for serious investment.

That said, there are some uniquely Chinese aspects worth noting:

1) The term “embodied intelligence”(具身智能) is now completely folded under the AI umbrella. At WAIC, nearly the entire second floor was taken over by robots—robot dogs and humanoids running around everywhere.

2) Chinese robot companies really shine in cost control thanks to local supply chains. For example, UBTECH explained that most of their core robot parts can be sourced one-stop in the Pearl River Delta, allowing them to price robots at about half the cost of SoftBank’s Pepper. One US tech observer even remarked that US robot prices may now average 30% higher than Chinese ones.

3) Robots are visibly entering public life. At both WAIC and WRC, I saw lots of kids—families and even tour groups—with children as young as 5 or 6 interacting with robots: playing soccer, boxing, etc. I’m not sure the same is happening in the US, but it gave me a strange sense of optimism about the future of robotics in China.

Once hardware issues — the body, joints, and locomotion — and balance control (the “brainstem” layer) are reasonably well solved, the real challenge becomes true embodied intelligence — the robot’s “cortex and cerebellum.” And that challenge is fundamentally tied to data scale.

The robotics field is nowhere near a “data wall.” On the contrary, it is in a state of severe data scarcity. Real-world robot data does not suffer from the sim-to-real gap, but it is extremely expensive and not scalable, and will inevitably hit a ceiling. Generalization is not inherently better because data is real; it depends on whether the data has sufficient coverage. If a model has never seen a particular scenario, it cannot generalize to it. The key question, therefore, is not real versus synthetic, but whether we can build a scalable data generation system.

After the Spring Festival Gala show, Wang Xingxing, CEO of Unitree, accepted an exclusive interview with CCTV News, and explained how the robots achieved this, and what globally first-of-its-kind technologies were unveiled.

Wang Xingxing told CCTV News, “We wanted the robots to push their limits on the Gala stage.” The robots performed three consecutive single-leg backflips with fluid continuity, drawing particular attention. Wang explained that the team had the robots leap from a custom-built launching platform, similar to how humans use assistive devices to accomplish extreme maneuvers. With the launcher, the robots could “jump” two to three meters high, complete front and side flips in midair, and land steadily. “Up close, it can almost jump as high as the ceiling.”

To realize this move, Wang and his team conducted hundreds of millions of training iterations in simulation before fine-tuning on physical robots. “This maneuver places extremely high demands on balance control, dynamic response, and landing stability. It’s a global first.”

Another breathtaking moment in WuBOT was the high-speed formation switching of more than twenty robots on stage. In last year’s YangBOT performance, robots changed formations by “walking slowly.” This year, they were able to execute rapid running transitions while weaving through formations and performing martial arts moves. This high-dynamic, high-coordination swarm control technology also made its global debut. “It’s a highly practical capability, laying the groundwork for future multi-robot coordination and single-robot dispatching in other real-world scenarios.”

Unitree team member Lian Yingying explained that to synchronize the movements of over twenty robots perfectly with the music, the martial arts performers, and the ever-changing stage terrain, the team meticulously fine-tuned each movement—sometimes refining timing to the tenth of a second.

From single-leg backflips to adapting to different surfaces, Wang said the goal was never simply to have robots perform a routine on flat ground. By pushing technological boundaries on the Gala stage, the team hopes to drive broader progress across the robotics industry. Director Chi Yuhan added that the program was conceived around the idea of a “future martial arts academy.”

The performance also included playful details. During the sparring segments, the robots were equipped with dexterous hands, allowing them not only to securely catch sticks but also to execute stylish stick-spinning techniques.

For the students of Henan Tagou Martial Arts School, it was their first time performing on stage with robots. Eleven-year-old Fu said sparring with robots felt novel and exhilarating. “It’s like training with senior disciples—it really boosts your fighting spirit.”

One of the most entertaining moments came in the latter half of the show: during a mock duel, a robot suddenly “collapsed drunkenly” to the ground, then sprang back up in a swift “carp flip.” Chi revealed that these moments were carefully scripted. During rehearsals, footage of a robot falling inspired the team to incorporate a playful “drunken boxing” flourish into the choreography.

Wang believes that showcasing cutting-edge robotics technology on the Spring Festival Gala stage to carry forward Chinese culture—especially kung fu—holds special meaning and marks a memorable moment globally. “Through the most advanced AI technology embodied in humanoid robots, Chinese kung fu can reach the world more effectively. In the coming years, robots combined with Chinese martial arts will surely captivate global audiences.”

“When cutting-edge AI is integrated into humanoid robots, they could one day become ‘fellow disciples’ training alongside children in a future martial arts academy,” Chi said. “We hope WuBOT plants a seed of technology in young minds.”

Why have robots perform? What can they do beyond the stage? In Wang’s view, the core of intelligent robotics is not flashy demonstrations, but ensuring that technology serves people. “Performing, running, martial arts—at their core, these activities improve stability, enabling robots to perform tasks that truly benefit everyday life.” Athletic capability is a prerequisite for intelligent robots: they must stand steadily and move reliably before they can work effectively. “If robots can execute martial arts routines amid complex formation shifts and rapid movement, it means they will be even more stable in future applications, giving people greater confidence.”

Wang also revealed that in 2026, several new products are in development, including more practically oriented service robots as well as exploratory models pushing technological frontiers.

“Enabling robots to genuinely advance human productivity—that is our shared goal, and it is what we are committed to pursuing this year.”

Wang Yi Meets with U.S. Secretary of State Rubio

Member of the Political Bureau of the CPC Central Committee and Foreign Minister Wang Yi met with U.S. Secretary of State Rubio on the sidelines of the Munich Security Conference.

Wang Yi stated that President Xi Jinping and President Donald Trump have provided strategic guidance for the development of China-U.S. relations. We should work together to implement the important consensus reached by the two heads of state and make 2026 a year in which China and the U.S. move toward mutual respect, peaceful coexistence, and win-win cooperation. Dialogue is better than confrontation, cooperation is better than conflict, and win-win is better than zero-sum. As long as both sides uphold the principles of equality, mutual respect, and reciprocity, they will be able to find ways to address each other’s concerns and properly manage differences. Both sides should make joint efforts to continuously expand the list of cooperation areas and reduce the list of issues, putting China-U.S. relations on a stable, healthy, and sustainable development track and sending more positive signals to the world.

Both sides agreed that the meeting was positive and constructive. They agreed to jointly implement the important consensus reached by the two heads of state, leverage political and diplomatic channels to coordinate efforts, support high-level exchanges between the two countries, enhance dialogue and cooperation in various fields, and promote stable development of China-U.S. relations.

The US readout

The below is attributable to Principal Deputy Spokesperson Tommy Pigott:

Secretary of State Marco Rubio met on February 13 with the Director of the Office of the CCP Central Foreign Affairs Commission and China’s Foreign Minister Wang Yi. The meeting was positive and constructive. Secretary Rubio emphasized the importance of results-oriented communication and cooperation on a range of bilateral, regional, and global issues.

They also discussed President Trump’s planned visit to China in April.

A few hours after the meeting between Wang Yi and U.S. Secretary of State Rubio, the U.S. Department of Defense released an updated 1260H list, adding 78 companies including Alibaba and BYD.

However, less than an hour later, the update disappeared. The latest notice on the Federal Register indicates that following the list’s publication, the Department of Defense received an interagency letter requesting its withdrawal. The document will remain publicly viewable until February 17, 2026.

This incident has drawn significant attention in China, with many trying to interpret the implications behind the move. Some analysts have noted that this scenario feels familiar: on May 13 of last year, the U.S. Department of Commerce issued guidance suggesting that procuring or using Huawei’s Ascend chips anywhere in the world would violate U.S. export controls. Two days later, BIS quietly revised the language, softening it to a warning that using Huawei chips “entails risks.” According to media reports, that change stemmed from a lack of coordination among U.S. government agencies. The rapid release and withdrawal of the updated 1260H list may be a recurrence of a similar interagency misalignment.

However, some speculations from the U.S. side suggest it might just be a technical error, that companies like YMTC and CXMT, which weren’t supposed to be removed from the list, were taken off by mistake. Now it’s just a correction, and the revised list might be reissued next week.

MOFCOM determined that under the framework of the EU’s Common Agricultural Policy (CAP), a range of measures—including direct payments, green subsidies, young farmer payments, redistributive payments, and support for agricultural processing investment—provided specific financial contributions to the dairy industry chain. Although these subsidies were primarily granted to upstream dairy farmers, the investigating authority found that the support did not remain confined to the agricultural segment. Instead, through raw milk pricing mechanisms, the benefits were transmitted downstream to processing enterprises, conferring a substantive benefit on dairy product manufacturers. In other words, although nominally framed as agricultural income support, the effects of the subsidies extended to exported dairy products.

With respect to injury determination, the ruling noted that during the investigation period, although domestic demand for relevant dairy products in China experienced some growth, the operating conditions of the domestic industry continued to deteriorate. Capacity utilization remained at a relatively low level; product prices declined; profits turned from positive to negative; inventories increased; cash flow came under pressure; and returns on investment became negative. In short, while the industry’s scale did not significantly contract, its profitability was clearly weakened. MOFCOM therefore concluded that the continued importation of subsidized EU products exerted price suppression and price depression effects on domestic products and was an important cause of the material injury suffered by the domestic industry. The three legal elements—subsidy, injury, and causal link—were all found to be satisfied.

At the enterprise level, the determination assigned specific duty rates to several sampled companies. Elle & Vire of France and its affiliated companies, FrieslandCampina of the Netherlands and Belgium and its affiliates, and Skånemejerier (Skånda) were selected as key sampled entities and were each assigned different subsidy rates. In addition, a number of European dairy companies participated in the investigation proceedings, including Arla of Denmark; BEL and Savencia of France; DMK of Germany; Milcobel of Belgium; and Lakeland of Ireland. These companies submitted questionnaire responses, attended hearings, or applied for price undertakings, and were therefore affected by the case to varying degrees.

From a broader perspective, the significance of this determination extends beyond a single product. At the legal level, it affirms an important logic: although EU agricultural subsidies are formally structured as income support to farmers, they may be recognized as indirect subsidies to downstream processed products and thus be subject to countervailing duties under the trade remedies framework.

This represents a direct challenge to the EU’s long-standing reliance on so-called “green box” or income-support measures within its agricultural policy. For the Chinese market, the ruling will raise the import costs of certain EU dairy products, potentially creating some room for price recovery for domestic producers, while at the same time possibly increasing cost pressures for downstream sectors such as high-end cheese, baking, and catering.

The above measure is under the background of the new round of tension in China-EU relations.

Not long ago, the European Commission formally introduced a revised draft of its Cybersecurity Act. Its core message can be summarized in one sentence: in critical sectors, equipment and components from “high-risk suppliers” must be gradually phased out. The key difference from the past is that what was previously framed as a recommendation not to use such suppliers has now become a legal requirement. Although no specific companies were named, it is widely understood that the term “high-risk suppliers” primarily alludes to Chinese enterprises. This development has drawn significant attention in China and is viewed as another negative move by the EU toward China.

At the same time, the EU has increasingly relied on its Foreign Subsidies Regulation (FSR) to initiate investigations against Chinese companies, and has upgraded probes into wind power and security screening equipment firms to in-depth investigations. In addition, the EU has intensively rolled out legislative proposals and implementing rules related to the Carbon Border Adjustment Mechanism (CBAM), including setting default carbon intensity values and planning to expand product coverage. From China’s perspective, the EU has disregarded the substantial progress China has made in green and low-carbon development, setting significantly elevated default carbon intensity benchmarks for Chinese products and planning to increase them annually over the next three years. The EU has also proposed legislation to expand the scope of CBAM from 2028 onward to cover approximately 180 downstream steel- and aluminum-intensive products, including machinery, automobiles and auto parts, and household appliances; and to amend the 2035 ban on new internal combustion engine vehicles, effectively relaxing certain green regulatory requirements within the EU.

China considers these measures to be clearly targeted and discriminatory, and a spokesperson for the Ministry of Commerce has expressed strong dissatisfaction with them.

Announcement No. 9 of 2026 of the Ministry of Commerce
Final Determination in the Countervailing Investigation on Certain Dairy Products Imported from the European Union

According to the Regulations of the People’s Republic of China on Countervailing Measures (hereinafter, the “Countervailing Regulations”), on August 21, 2024, the Ministry of Commerce (the “investigating authority”) issued Announcement No. 34 of 2024, deciding to initiate a countervailing investigation into certain dairy products imported from the European Union (hereinafter, the “products under investigation”).

The investigating authority examined whether the products under investigation were subsidized and the amount of any subsidies; whether the products under investigation caused injury to the domestic industry and the degree of such injury; and the causal link between the subsidies and the injury. Based on the findings and pursuant to Article 25 of the Countervailing Regulations, the investigating authority issued a preliminary determination on December 22, 2025, finding that the products under investigation were subsidized, that the relevant domestic dairy industry in China had suffered material injury, and that there was a causal relationship between the subsidies and the material injury.

Following the preliminary determination, the investigating authority continued its investigation into the subsidies and subsidy amounts, injury and the degree of injury, and the causal relationship between the subsidies and the injury. The investigation has now concluded. Based on the findings, and pursuant to Article 26 of the Countervailing Regulations, the investigating authority has made a final determination (see Annex 1). The relevant matters are hereby announced as follows:

I. Final Determination

The investigating authority has finally determined that certain dairy products imported from the European Union are subsidized, that the relevant domestic dairy industry in China has suffered material injury, and that there is a causal relationship between the subsidies and the material injury.

II. Countervailing Measures

Pursuant to Article 39 of the Countervailing Regulations, the Ministry of Commerce has submitted to the Customs Tariff Commission of the State Council a recommendation to impose countervailing duties on certain dairy products imported from the European Union. The Customs Tariff Commission of the State Council has decided, based on the Ministry of Commerce’s recommendation, that countervailing duties will be imposed on certain dairy products imported from the European Union as of February 13, 2026.

The products under investigation are described as follows:

Scope of investigation: Certain dairy products imported from the European Union.

Name of products under investigation: Certain dairy products.

English name: Certain dairy products.

Product description: The products under investigation specifically include fresh cheese (including whey cheese) and curd; processed cheese (whether or not grated or powdered); blue-veined cheese and other cheeses with veining produced by Penicillium roqueforti; other cheeses not elsewhere specified; and milk and cream, not concentrated and not containing added sugar or other sweetening matter, with a fat content by weight exceeding 10%.

Main use: Mainly for direct human consumption as food, either as-is or after processing.

HS codes under the Customs Import and Export Tariff of the People’s Republic of China: 04015000, 04061000, 04062000, 04063000, 04064000, 04069000.

The countervailing duty rates applicable to EU companies are set out in Annex 2 to this Announcement.

III. Method of Collection of Countervailing Duties

As of February 13, 2026, importers, when importing certain dairy products originating in the European Union, shall pay the corresponding countervailing duties to China Customs. Countervailing duties shall be levied ad valorem on the dutiable value of the imported goods as determined by Customs, calculated as follows:

Countervailing duty payable = dutiable value as determined by Customs × countervailing duty rate.

Import VAT shall be levied ad valorem on the taxable base consisting of the dutiable value as determined by Customs plus customs duties and countervailing duties.

IV. Retroactive Collection of Countervailing Duties

With respect to the provisional countervailing duty deposits provided by relevant importers to China Customs pursuant to the preliminary determination for imports made from December 23, 2025 to February 12, 2026, countervailing duties shall be assessed and converted from such deposits based on the product scope and countervailing duty rates as determined in the final determination, and import VAT shall be assessed at the applicable VAT rate. Any portion of the provisional deposits exceeding the countervailing duties, as well as any resulting overpaid import VAT, shall be refunded by Customs; any shortfall shall not be collected.

No countervailing duties shall be collected retroactively on certain dairy products originating in the European Union that were imported prior to the implementation of the provisional countervailing measures.

V. Duration of Collection of Countervailing Duties

The countervailing duties on certain dairy products imported from the European Union shall be imposed for a period of five (5) years starting from February 13, 2026.

VI. Review

During the period in which countervailing duties are imposed, interested parties may apply to the investigating authority for a review in accordance with the relevant provisions of the Countervailing Regulations.

VII. Administrative Reconsideration and Administrative Litigation

If any party is dissatisfied with the final determination in this case and the decision to impose countervailing duties, it may, pursuant to Article 52 of the Countervailing Regulations, apply for administrative reconsideration in accordance with law, or file an administrative lawsuit with a people’s court in accordance with law.

VIII. Effective Date

This Announcement shall take effect as of February 13, 2026.

According to a readout released by the Department of Arms Control of the Chinese Ministry of Foreign Affairs, Li elaborated on China’s concept of “human-centered military artificial intelligence” during the summit.

All countries should uphold a bottom line of prudence and responsibility, abandon the obsession with absolute military advantage, and earnestly safeguard strategic balance and stability; adhere to a people-centered approach, comply with international humanitarian law, and ensure that relevant weapon systems always remain under human control; follow the principle of “AI for good,” and promote military applications of artificial intelligence that serve the maintenance of peace and security; implement the principle of agile governance by balancing security controls, technological development, and peaceful use through tiered and categorized management; and uphold multilateralism, support the United Nations in playing its due role, and promote the establishment of governance frameworks and standards based on broad consensus.

Military intelligentization is a major trend in the development of armed forces worldwide. How to responsibly use artificial intelligence in the military domain concerns the shared future of all humanity and constitutes a common challenge of our time. China maintains that the international community should embrace a global governance philosophy of extensive consultation, joint contribution, and shared benefits, as well as a security concept that is common, comprehensive, cooperative, and sustainable, and work together to build effective governance mechanisms to ensure that artificial intelligence always develops in a direction conducive to the progress of human civilization.

As a major artificial intelligence power, China has consistently attached great importance to risk prevention and security governance in the military application of AI, and has adhered to the concept of “human-centered military artificial intelligence. Countries, when applying relevant technologies, should uphold a bottom line of prudence and responsibility, maintain a people-centered approach, follow the principle of AI for good, implement agile governance, adhere to multilateralism, and promote military AI applications that serve the maintenance of peace and security.

China has incorporated “strengthening AI governance” into the recommendations for its 15th Five-Year Plan. In accordance with global trends in military transformation, objective national security needs, and the practical requirements of building national defense and armed forces, China will advance military intelligentization, refrain from engaging in an AI arms race with other countries, and maintain policy transparency.

China is also actively leveraging the advantages of military AI in peacekeeping and humanitarian assistance, with positive results. Li stressed that China will continue to uphold the principles of openness, inclusiveness, and mutual learning, strengthen communication and exchanges with all countries, deepen practical cooperation, and jointly contribute to global governance of military artificial intelligence.

Beijing Daily reporter Jin Liang wrote that 85 countries participated in the summit. After two days of talks, about one-third of the participating countries issued a joint declaration on regulating the deployment of AI technologies in warfare, but only 35 countries ultimately signed the declaration; neither China nor the United States signed.

According to Jin, the United States’ refusal to sign binding international rules on military AI stems from multiple strategic considerations. At the core is Washington’s concern that international rules could constrain rapid technological iteration and military deployment flexibility, thereby eroding the first-mover advantage and technological gap it has already established in the field of military AI.

At the same time, the United States is highly sensitive about rule-making authority and is unwilling to co-shape standards with non-“like-minded” countries such as China within multilateral frameworks like the United Nations. Instead, it prefers to build exclusive “small-circle” governance rules within its alliance system to preserve its technological and discursive dominance.

In addition, by avoiding explicit international obligations, the United States retains strategic ambiguity in sensitive areas such as autonomous weapons and battlefield AI decision-making, avoids transparency constraints, and preserves room for “gray-zone” military operations.

Finally, at the level of public discourse, the United States avoids substantive constraints while portraying itself as a “responsible user of AI,” shifting risks and pressure onto competitors and using the banner of rules to serve the practical objective of containing rivals and consolidating strategic advantages.

As for why China did not sign the declaration, Jin argues that the main reasons lie in doubts about the vague formulation of principles such as “responsible use” and the lack of mechanisms to balance the technological advantages of early-mover states. China believes that the existing framework could entrench Western-dominated technological hegemony and constrain the technological autonomy and security space of developing countries.

China has long advocated multilateral governance, emphasizing that international rules should balance security and development, and opposing the politicization of technology or bloc confrontation. Certain provisions of the declaration are seen as potentially embedding ideological bias and failing to reflect fairness.

At the same time, China is rapidly catching up in the field of military AI and needs to preserve strategic space for independent research, development, and deployment, avoid being constrained by unreasonable rules, and ensure that national sovereignty and security are not subject to external interference.

Jin further notes that two shared structural obstacles prevented both China and the United States from signing the declaration.

First, military AI involves highly sensitive defense secrets and operational capabilities, making verification and enforcement of any international rules extremely difficult. While countries seek to mitigate risks through rules, they also worry that adversaries might “free-ride” or violate commitments, creating a classic prisoner’s dilemma.

Second, the pace of AI technological iteration far exceeds the cycle of rule-making. The declaration’s largely principled provisions are insufficient to address concrete risks such as autonomous weapons, algorithmic bias, and battlefield misjudgment. Both China and the United States therefore view the declaration as “incomplete,” lacking practical binding force, and of limited value to sign.