China spy chief warns AI threatens party rule
Archived — this story has rotated out of today’s deck. It is kept here in full.
The gist
China's spy chief Chen Yixin warned AI threatens the party's political and ideological security.
The warning draws a red line before US-China AI safety talks.
Background
Chen Yixin heads China's Ministry of State Security, and his article in the state-run journal China Cyberspace frames AI as a national security matter rather than a shared humanitarian one. He cited next-generation US-led models such as Anthropic's Claude Mythos and OpenAI's GPT-5.5-Cyber as new cyber threats, and said hostile forces use generative AI to fabricate political rumours and wage cognitive warfare against China. The piece lands as Beijing seeks a leading role in global AI governance while criticising US restrictions, and ahead of expected US-China talks on AI safety risks.
How it unfolded
- Jul 2026At the World Artificial Intelligence Conference in Shanghai, Xi Jinping said AI should 'always remain under human control' and that authorities should watch both intrinsic and derivative risks.
- Sep 10, 2026Anthropic published a threat intelligence report saying it disrupted and banned accounts linked to China-based actors, including operations using Claude for public opinion monitoring and stability maintenance surveillance.
- Sep 13, 2026Chen Yixin's article was published on the social media account of China Cyberspace, a journal run by the Cyberspace Administration of China.
- Sep 14, 2026Chen Yixin published the article warning AI could threaten the party's political, institutional and ideological security, calling for more party control over AI and stricter oversight.
- Sep 15, 2026China rejected US calls to curb artificial intelligence as 'fearmongering'.
- Sep 24, 2026A meeting between President Trump and Xi Jinping is expected in Washington, with US-China talks on AI safety risks expected ahead of it.
Who’s saying what
- Official
- Chen Yixin said hostile forces abuse generative AI to fabricate political rumours and spread harmful information, launching 'public opinion wars and cognitive warfare' that directly threaten China's political, institutional and ideological security.
- Analysts
- Henry Gao of Singapore Management University said Beijing views AI safety through an adversarial lens and is signalling that 'data sovereignty and political security will never be traded away for international safety accords'.
- Expert
- Geoffrey Hinton backed Anthropic CEO Dario Amodei's call to slow AI development, saying practically nobody knows if AI can 'be kept under control'; Demis Hassabis said the direction is correct though details need refining.
- Caution
- China rejected US calls to curb artificial intelligence as 'fearmongering'.