Chinese President Xi Jinping called for deeper international coordination on artificial intelligence development and governance on Friday, arguing that the technology should not be controlled by any one country and urging governments to establish safeguards that keep AI under human oversight.
Speaking at the opening of the World Artificial Intelligence Conference in Shanghai, Xi presented cooperation as an alternative to national competition over increasingly capable AI systems. France 24, reporting with AFP, said his remarks addressed concerns ranging from cybersecurity and armed conflict to employment and the global economy.
Xi warned against applying national security rationales too broadly in the AI sector or allowing one country’s security priorities to override those of others. His intervention comes amid continuing restrictions by the United States and the European Union on some Chinese technology imports, measures that have been justified on security grounds. It also arrives as policymakers worldwide debate who should control access to advanced models, computing capacity and the chips needed to build powerful AI systems.
The Chinese leader called for laws and regulations alongside technical monitoring, early-warning mechanisms and emergency-response systems. According to France 24, Xi said these measures were needed to ensure that AI remains under human control and advocated a people-centered approach to the technology.
The speech places China’s push for international AI rules alongside its ambition to become a leading producer and exporter of the technology. Chinese developers have been narrowing parts of the performance gap with prominent US laboratories, while lower prices and customizable open-source models have helped some Chinese systems attract users overseas. That combination gives Beijing both a commercial interest and a diplomatic stake in shaping emerging global standards.
Governance has become an increasingly urgent issue as general-purpose AI models gain the ability to generate software, analyze large volumes of information and operate other digital tools. Officials and researchers have raised concerns about possible military applications, cyberattacks and the use of AI by terrorist groups. The same capabilities can support legitimate research and commercial activity, making legal controls difficult to design without limiting beneficial uses or concentrating access among a small number of governments and companies.
Xi’s comments followed the establishment of a new intergovernmental group intended to support AI consultation and cooperation. France 24 reported that Chinese Foreign Minister Wang Yi and representatives of 29 countries, including Russia, Pakistan and Indonesia, signed an agreement on Thursday creating the World Artificial Intelligence Cooperation Organization.
State media described the Shanghai-headquartered body as a forum for promoting collaboration and the orderly development of AI. The report did not detail what legal authority the organization would hold, how its decisions would be made or whether its standards would be binding. Those questions will be important in determining whether it develops into a rule-setting institution or primarily serves as a venue for diplomatic coordination.
International AI governance remains fragmented. National and regional authorities are developing their own approaches to safety testing, data protection, intellectual property, competition and accountability, while global forums seek areas of consensus. Differences over national security, access to advanced semiconductors and the openness of model technology complicate efforts to establish common rules.
China’s proposal also reflects a broader contest over the values and institutions that will govern AI. Rules covering model evaluations, incident reporting, human oversight and the deployment of autonomous systems can influence which companies gain market access and how products are designed. Standards framed as technical requirements can therefore carry significant commercial and geopolitical consequences.
The four-day Shanghai conference brings together more than 1,000 Chinese technology companies as well as government officials, researchers and industry representatives, according to France 24. About 3,000 products are being displayed, spanning semiconductor systems for AI computing and consumer devices capable of operating applications autonomously.
Poe Zhao of the analysis publication Hello China Tech described the conference to AFP as a key annual indicator of the Chinese AI industry’s direction. Zhao said the United States continued to lead in advanced chips, frontier computing infrastructure and highly capital-intensive model development, while identifying China as its nearest broad-based competitor.
Products highlighted at the conference illustrate a shift from standalone chatbots toward systems capable of taking actions. AI agents can converse with users, manage software and carry out multistep tasks. Their movement into everyday devices and workplace systems creates additional legal questions because errors may have direct consequences beyond the production of inaccurate text or images.
When an AI system can operate an application, make a recommendation or execute part of a business process, regulators and courts may need to determine how responsibility is divided among developers, deployers and users. Effective oversight can also require records showing how a system acted, what permissions it had and when a human intervened. Xi’s emphasis on monitoring and emergency response addresses part of that challenge at a high level, although the speech as reported did not set out detailed compliance mechanisms.
The event also featured new Chinese models and computing products. France 24 reported that Beijing startup Moonshot AI released its Kimi K3 flagship model early Friday, with reported performance intended to compete with leading US offerings. Other showcased technology included MiniMax’s M3 model, a mass-produced phone featuring an autonomous AI agent and Huawei’s Atlas 950 computing architecture.
China has made AI a central element of its industrial policy, supporting an ecosystem that extends from semiconductor production to consumer applications. State media, citing officials, said daily Chinese consumption of tokens—the units commonly used to measure model processing—had increased a thousandfold in two years. Official data cited in the report valued China’s AI market at 1.2 trillion yuan, or about $177 billion, in 2025 and projected growth of more than 30 percent this year.
Intellectual property is another major component of the governance debate. France 24 cited World Intellectual Property Organization data showing that China recorded more than 43,000 generative AI patent filings in 2024 and 2025, more than any other country. Patent volume does not by itself establish technological leadership, but it demonstrates the scale of legal and commercial activity surrounding Chinese AI development.
Generative AI has also produced disputes over the material used to train models and the ownership of machine-generated output. Governments are weighing how existing copyright, patent and trade-secret rules apply to systems trained on vast datasets. The growing international adoption of Chinese open-source models adds a cross-border dimension, because model licenses, data rules and liability requirements can differ among jurisdictions.
Xi’s appeal for cooperation therefore extends beyond diplomatic language. Any durable international framework would have to reconcile competing positions on security restrictions, open-source development, commercial access and state oversight. The Shanghai initiative offers China a platform to advance its preferred approach, but its practical influence will depend on participation, transparency and whether governments can translate broad commitments into compatible legal and technical standards.
Sources: Generative AI intellectual property