China hits back at Anthropic CEO over AI "fearmongering" claim

China's government has pushed back hard against comments from the head of Anthropic, the company behind the Claude AI models, calling his warnings about Chinese artificial intelligence "fearmongering" that could damage global cooperation.

The dispute started on 12 September, when Anthropic CEO Dario Amodei published an essay called "We Must Pace the Frontier". In it, he argued the United States should keep its restrictions on selling advanced computer chips to China, and proposed banning a technique called "distillation" — where one company trains its AI on the outputs of another company's model, effectively learning from it without building everything from scratch. Amodei warned that a "Chinese lead in AI would pose grave danger for the United States and the world."

China's Foreign Ministry responded at its regular press conference on Monday 14 September. Spokesperson Guo Jiakun said: "Fearmongering, confrontation and vicious competition will only disrupt the process of global AI governance -- which serves no one's interest." He called for countries to work together on AI instead.

State-run newspaper the Global Times had already gone further, describing Amodei's essay as "packed with containment provisions targeting China" and calling it "in essence, a 'Cold War playbook' for the AI sector," according to the Associated Press.

The row did not come out of nowhere. Days earlier, three US agencies — the FBI, NSA and CISA — issued a joint advisory accusing Chinese AI developers of "aggressive, malicious" attempts to distil capabilities from Claude and from OpenAI's GPT models. China's Commerce Ministry called those accusations groundless. Separately, Anthropic published a threat report on 10 September naming seven Chinese labs it said were linked to fraudulent accounts on its platform.

If you use AI tools built in the US or China, this dispute is part of a wider fight over who controls the chips and techniques that power those tools. Export controls on advanced chips could affect how fast AI models improve in different countries, and restrictions on distillation could change how new AI companies are able to build competing products.

Presidents Trump and Xi were expected to discuss AI governance when they met on 24 September. Trump had separately criticised Amodei's calls for safety guardrails on AI development, according to The Korea Times.