China rejects Anthropic chief's call to slow AI race

Anthropic chief executive Dario Amodei published a post over the weekend urging a "global pace" for advanced model development, but the condition he attached — preserving the widest possible lead for democracies over autocracies — rendered the proposal a non-starter in Beijing. Amodei called on the United States to tighten enforcement against chip smuggling and to block model distillation by Chinese companies, moves designed to freeze the rival in place rather than slow both sides together.

The condition that breaks the deal

Kevin Xu, founder of Interconnected Capital and a former Obama White House staffer, says China's AI safety community shares many of the same concerns as US labs. The gap is procedural, not principled: Anthropic's vision requires China's submission and constraint first to gain leverage, while Beijing will engage only from a position of equality. Alvin Wang Graylin, a senior fellow at Asia Society and a Stanford researcher, notes that Amodei's own essay ties the pacing to widening the gap, making the invitation to cooperate read as inauthentic to Chinese readers.

An admission of no strategy

Amodei argued that weakening China's AI industry would actually increase US leverage and make a future agreement more likely, yet he did not explain how that would lure Beijing to the negotiating table. In a subsequent television interview, the chief executive simply admitted he has no idea how to negotiate with China, framing the entire initiative as a thought experiment rather than actionable policy.

A sharp official response

Chinese foreign ministry spokesperson Guo Jiakun did not wait. At a Monday press conference he labelled the proposal "intimidation, confrontation and vicious competition" that would only damage global cooperation. The next day the ministry stressed that China places equal weight on development and security. The phrasing exposes the fundamental divide: where Silicon Valley speaks of existential risks such as a rogue agent destroying humanity, the Chinese leadership's primary concern is domestic social stability and the Communist Party's grip on power.

Scepticism about AGI itself

Underlying the exchange is a broader Chinese stance: policymakers in Beijing struggle to take slowdown calls based on imminent AGI seriously because they doubt the feasibility and timelines of artificial general intelligence. When the founding premises diverge — existential risk versus regime stability — the framework for cooperation collapses before it can be built.