China's spy chief flags AI threat, warns of 'cognitive warfare' & threats to security

China's Ministry of State Security has issued its first public warning about the national security risks posed by artificial intelligence. Minister Chen Yixin accused foreign actors of using generative AI to spread political rumors, manipulate public opinion and threaten critical infrastructure through "public opinion and cognitive warfare".

These activities, Chen wrote, directly threaten China's political, institutional and ideological security. AI could also compromise China's cyber defenses, sensitive data and military systems, Chen stressed. The spokesperson cited Anthropic's Claude Mythos and OpenAI's ChatGPT-5.5-Cyber as systems that could exploit China's critical information infrastructure.

Foreign intelligence agencies are already using AI-enabled tools to gather sensitive data, Chen alleged. The warning comes as China and the US prepare for AI governance talks. US and Chinese presidents are expected to discuss the issue at a September 24 meeting. US tech executives have expressed concern over the pace of AI development, with Anthropic CEO Dario Amodei urging the industry to slow advancements.

China's foreign ministry dismissed calls for confrontation over AI, stating that fearmongering and competition would disrupt global AI governance.