China's Minister for State Security, Chen Yixin, has publicly warned that artificial intelligence poses risks to the Communist Party's grip on national security and social governance, in an article for China Cyberspace Magazine. He frames AI as the central arena of great-power competition, citing fears that foreign AI tools could be weaponised to exploit software vulnerabilities, attack infrastructure or steal state secrets, while also flagging concerns about algorithmic "black boxes," data misuse and weakened accountability eroding public trust and social order. This matters because it signals Beijing is likely to tighten control over AI development and foreign technology access rather than relax it.

Chen singled out products such as OpenClaw for allegedly allowing remote device control and leaking sensitive user data, and blamed domestic users' poor security awareness for data leaking overseas via foreign AI tools. His prescribed remedy is for China to pursue "technological sovereignty," modernise its national security apparatus per Xi Jinping's guidance, and introduce dedicated AI laws covering the full chain from research to accountability — a stance that dims prospects for Nvidia and AMD selling GPUs into China. A day later, the Cyberspace Administration of China released version 3.0 of its AI Safety Governance Framework, backing regulatory sandboxes and risk controls, a notably more cautious approach than that of US President Donald Trump, who this week dismissed AI safety concerns as a "hoax."