China warns that AI development could threaten national security, escalating global policy debate over AI regulation and export controls amid US-China tech rivalry.

China has issued one of its sternest warnings regarding artificial intelligence, with the country’s top intelligence official cautioning that the technology poses potential threats to political stability, critical infrastructure, and national security. Minister of State Security Chen Yixin warned that foreign actors could leverage generative AI, deepfakes, and automated information campaigns to manipulate public opinion and conduct what Beijing terms “cognitive warfare.” This statement represents the first major public intervention on AI from the Ministry of State Security, explicitly anchoring national security considerations within Beijing’s rapidly evolving regulatory framework for the technology.

Chen’s assessment diverges sharply from prevailing narratives in Silicon Valley. While industry leaders such as Anthropic chief executive Dario Amodei have increasingly emphasized the risks of highly capable AI systems operating outside human control or being weaponized to develop hazardous technologies, Beijing places primary emphasis on safeguarding political and social stability. Chen cautioned that AI could be deployed to manufacture misinformation, exploit cybersecurity vulnerabilities, and target critical infrastructure, while large-scale automated influence operations could erode China’s political and ideological security. These concerns signal the prospect of heightened state oversight over China’s fast-expanding AI ecosystem, which includes Alibaba, DeepSeek, ByteDance, Baidu, and Moonshot AI. The warning arrives as Chinese regulators intensify scrutiny of AI-generated content; the Cyberspace Administration of China has been actively policing fake information, impersonation, and other misuses of generative AI across domestic online platforms.

Simultaneously, Beijing is vigorously contesting efforts to frame its own advancement in cutting-edge AI as a global security liability. China has publicly criticized Amodei’s recent proposal that the United States should sustain restrictions on exporting advanced AI chips to China to preserve a technological edge. Beijing dismissed the argument as fearmongering and instead advocated for multilateral cooperation on AI governance. This dynamic places China in a strategically complex position: issuing domestic alerts about AI-related national security threats while actively resisting U.S. attempts to leverage those same concerns to justify limiting China’s access to advanced computing hardware.

The friction is also expanding beyond semiconductor export controls. U.S. authorities recently alleged that Chinese AI firms are utilizing advanced American models to accelerate their own development through a technique known as model “distillation,” claims Beijing firmly rejected as an effort to entrench U.S. industry dominance. The escalating tensions further cement AI’s status alongside semiconductors and quantum computing as a core arena of competition between the world’s two largest economies. As Chinese developers continue to release increasingly sophisticated models despite constraints on accessing the most advanced U.S. chips, Beijing’s latest warning indicates that the subsequent phase of the AI race will not hinge solely on which entity constructs the most powerful system. Rather, the capacity to govern how these models are deployed—and who retains control over the foundational technology—is rapidly emerging as an equally decisive factor.