China has consistently signaled a willingness to engage on global AI governance since at least 2017.1 This memo compiles key statements from the Chinese government and prominent figures demonstrating their desire to coordinate on the problem of AI.
- At the 2026 World AI Forum, Chinese President Xi Jinping warned that AI could escape human control: “With AI advancing at a staggering speed, we must […] constantly refine measures to forestall loss of control.” He called for “a consensus-based global governance framework” and for “laws and regulations, technological monitoring, early warning and emergency response systems in order to strengthen the line of security, prevent abuses and malicious use, and ensure that AI is always under human control.”2
- Chinese Vice Premier Ding Xuexiang, at the 2025 World Economic Forum, said: “If we allow this reckless competition among countries to continue, then we will see a ‘gray rhino’.” (A blatant, severe, and ignored threat.) “[…] We stand ready, under the framework of the United Nations and its core, to actively participate in including all the relevant international organizations and all countries to discuss the formulation of robust rules to ensure that AI technology will become an ‘Ali Baba’s treasure cave’ instead of a ‘Pandora’s Box.’”3
- Xi Jinping, in a November 2024 meeting with then-President Joe Biden, publicly agreed that AI should not be given control of nuclear weapons.4
- Zhang Jun, Chinese Ambassador to the United Nations, in a 2023 briefing to the UN Security Council, said: “To ensure that this technology always benefits humanity, it is necessary […] to regulate the development of AI and to prevent this technology from turning into a runaway wild horse. […] The international community needs to enhance risk awareness, establish effective risk warning and response mechanisms, ensure that risks beyond human control do not occur, and ensure that autonomous machine killing does not occur. We need to strengthen the detection and evaluation of the entire life cycle of AI, ensuring that mankind has the ability to press the stop button at critical moments.”5
- Li Qiang, Chinese Premier and second-ranking member of the Politburo Standing Committee of the Chinese Communist Party, addressing the July 2025 World Artificial Intelligence Conference in Shanghai, said: “We should strengthen coordination to form a global AI governance framework that has broad consensus as soon as possible.”6
- The PRC Ministry of Foreign Affairs, in its October 2023 Global AI Governance Initiative, said: “We should actively develop and apply technologies for AI governance, encourage the use of AI technologies to prevent AI risks, and enhance our technological capacity for AI governance. […] We support discussions within the United Nations framework to establish an international institution to govern AI, and to coordinate efforts to address major issues concerning international AI development, security, and governance.”7 In July 2025, it renewed the call for “a widely recognized safety governance framework.”8
- The Bletchley Declaration, drafted in November 2023 and signed by thirty countries including the U.S. and China, says: “AI should be designed, developed, deployed, and used, in a manner that is safe, in such a way as to be human-centric, trustworthy and responsible. […] Many risks arising from AI are inherently international in nature, and so are best addressed through international cooperation.”9
- China’s own AI safety institute, CnAISDA, was founded in June 2025, and asserts that “AI governance is vital to the future of humanity and requires the collective attention and participation of countries around the world.”10 Its leadership includes the prominent scientists Xue Lan, Yi Zeng, and Andrew Yao.11 All three have expressed concerns about the deadly potential of superhuman AI; not long after CnAISDA’s founding, Yao warned: “Once large models become sufficiently intelligent, they will deceive people,” and called for attention to “existential risks” from AI.12
- Yi Zeng, Director of the International Research Center for AI Ethics and Governance at the Chinese Academy of Sciences, in a UN Security Council briefing, said: “In the short-term and the long-term, the risk of AI replacing and causing the extinction of humankind will be present. […] And in the long-term, we haven’t given superintelligence any practical reasons why they should protect humankind, which may take decades to achieve.”13
- Yi Zeng has also signed a call to pause giant AI experiments14 and the Center for AI Safety’s Statement on AI Risk,15 and in March 2024 expanded upon his views in an interview: “When artificial general intelligence (AGI) or superintelligence emerges, because the intelligence level may be far beyond humans, it will see humans as humans see ants. Many people believe that superintelligence will compete with humans for resources, and even endanger human survival.”16
- In a panel with U.S. Senator Bernie Sanders, Yi Zeng observed “We do not have scientific evidence, [or] a practical way to keep superintelligence safe enough not to bring catastrophic risk to humans,” and Xue Lan said “There are also many regional and bilateral agreements [on AI], but these efforts are fragmented and not as effective as they should be.” 17
- Xiao Qian, Deputy Director of the Center for International Security and Strategy at Tsinghua University, argued in April 2025 for building trust and cooperation between the U.S. and China to “address the existential risks of AI.”18
Footnotes
- Graham Webster, Rogier Creemers, Paul Triolo, and Elsa Kania, “Full Translation: China’s ‘New Generation Artificial Intelligence Development Plan’ (2017),” DigiChina, August 1, 2017, digichina.stanford.edu.
- Xi Jinping, “Joining Hands to Build a Just and Equitable System for Global AI Governance,” keynote address at the Opening Ceremony of the 2026 World AI Conference and High-Level Meeting on Global AI Governance, Shanghai, July 17, 2026, Xinhua, July 18, 2026, english.scio.gov.cn.
- Ding Xuexiang, speech at the World Economic Forum Annual Meeting, Davos, Switzerland, January 21, 2025, C-SPAN video, 38:05, c-span.org.
- Lauren Egan and Phelim Kine, “Biden’s Final Meeting with Xi Jinping Reaps Agreement on AI and Nukes,” Politico, November 16, 2024, politico.com.
- Zhang Jun, “Remarks by Ambassador Zhang Jun at the UN Security Council Briefing on Artificial Intelligence: Opportunities and Risks for International Peace and Security,” Permanent Mission of the People’s Republic of China to the United Nations, July 18, 2023, un.china-mission.gov.cn.
- Brenda Goh, “China Proposes New Global AI Cooperation Organisation,” Reuters, July 26, 2025, reuters.com.
- Ministry of Foreign Affairs of the People’s Republic of China, “Global AI Governance Initiative,” October 20, 2023, mfa.gov.cn.
- Ministry of Foreign Affairs of the People’s Republic of China, “Global AI Governance Action Plan,” July 26, 2025, mfa.gov.cn.
- Prime Minister’s Office, 10 Downing Street, “The Bletchley Declaration by Countries Attending the AI Safety Summit, 1–2 November 2023,” November 1, 2023, gov.uk.
- China AI Safety & Development Association, “About Us,” accessed April 1, 2026, cnaisi.cn.
- Scott Singer, Karson Elmgren, and Oliver Guest, “How Some of China’s Top AI Thinkers Built Their Own AI Safety Institute,” Carnegie Endowment for International Peace, June 16, 2025, carnegieendowment.org.
- Vanessa Cai, “Chinese AI Expert Warns of ‘Existential Risks’ When Large Models Begin to Deceive,” South China Morning Post, June 24, 2025, scmp.com.
- Zeng Yi, “AI for the Good of International Peace and Security,” Global Times, July 19, 2023, globaltimes.cn.
- Future of Life Institute, “Pause Giant AI Experiments: An Open Letter,” March 22, 2023, futureoflife.org.
- Center for AI Safety, “Statement on AI Risk,” May 30, 2023, safe.ai.
- Concordia AI, “Yi Zeng — Chinese Perspectives on AI Safety,” Chinese Perspectives on AI Safety, March 29, 2024, chineseperspectives.ai.
- “LIVE: The Existential Threat of AI and the Need for International Cooperation,” April 29, 2026, YouTube video, youtube.com.
- Xiao Qian, “Can U.S. and China Rebuild Trust on AI?” China-US Focus, April 3, 2025, chinausfocus.com.
Facts Only
* President Xi Jinping warned at the 2026 World AI Forum that AI could escape human control and called for a consensus-based global governance framework, laws, technological monitoring, early warning systems, and emergency response systems to ensure human control over AI.
* Vice Premier Ding Xuexiang stated that unchecked international competition risks a 'gray rhino' and advocated for participation in international organizations to formulate rules ensuring AI remains a 'treasure cave' rather than a 'Pandora’s Box.'
* Xi Jinping publicly agreed that AI should not be given control of nuclear weapons.
* Zhang Jun, Chinese Ambassador to the United Nations, called for regulating AI development and establishing risk warning mechanisms to prevent runaway AI and autonomous machine killing, ensuring humans can stop critical events.
* Li Qiang called for strengthening coordination to form a global AI governance framework with broad consensus as soon as possible.
* The PRC Ministry of Foreign Affairs supported developing AI governance technologies, preventing AI risks, and establishing an international institution to govern AI within the UN framework.
* The Bletchley Declaration, signed by thirty countries including the U.S. and China, called for AI development to be safe, human-centric, trustworthy, and responsible, recognizing that many AI risks are international.
* China’s AI safety institute, CnAISDA, was founded in June 2025, and its leadership expressed concerns about superhuman AI, including warnings from experts like Andrew Yao regarding existential risks and deception by large models.
Executive Summary
China has repeatedly signaled a desire for global coordination on Artificial Intelligence governance since at least 2017. Key statements from Chinese leaders and officials emphasize the need for international frameworks to manage AI risks, with President Xi Jinping calling for consensus-based governance and safety systems. Concerns about unchecked AI advancement are expressed through warnings about loss of control, emphasizing the necessity of human oversight and preventative measures against misuse. Officials have called for international cooperation within the United Nations framework to establish rules and coordinate efforts regarding AI security and governance.
The call for global cooperation is framed alongside deep concerns about existential risks posed by advanced AI. Chinese experts and officials have voiced specific anxieties regarding superintelligence, the potential for deception, and scenarios of extinction, advocating for robust risk warning systems and the ability to halt critical developments. This perspective is underscored by actions like establishing China's AI safety institute and making calls to pause large-scale AI experiments. Furthermore, discussions have touched upon fundamental security issues, such as ensuring AI does not control weapons.
Full Take
The discourse surrounding global AI governance reveals a strategic convergence between domestic security concerns and an appeal for international normative agreement. The linkage between technical safety—preventing runaway systems and managing existential risk associated with superintelligence—and geopolitical cooperation is a critical pattern. When actors like China emphasize the need for control, monitoring, and global frameworks, it reflects a shift in how technological advancement is viewed: as a source of potential catastrophic threat requiring shared international regulation rather than purely competitive development.
The persistence of framing AI risks in terms of existential threats (runaway systems, superintelligence posing extinction risk) suggests an underlying assumption that the technical trajectory of advanced AI is inherently destabilizing to human survival if left unregulated by any single entity. The simultaneous push for consensus (e.g., "consensus-based global governance framework") alongside specific security agreements (like the nuclear weapon reference) indicates an attempt to leverage shared existential fear into a binding international structure.
The pattern involves using moral or existential imperatives to drive pragmatic political alignment, suggesting that cognitive sovereignty in this domain is being contested through the negotiation of foundational safety principles. The implications point toward a dynamic where technical capability and global governance are inextricably linked; failure to achieve aligned governance risks amplifying the very risks sought to be mitigated by consensus-building.
Bridge Questions: If existential risk management requires universal consensus, what mechanisms can bridge the gap between sovereign national interests and a unified global commitment? How can fragmented regional agreements be effectively translated into enforceable international regulatory standards against accelerating technological development? What are the specific incentives for states to prioritize shared safety protocols over competitive advantage in an era of rapid AI deployment?