The Claim, in Two Parts

The claim under review has two components. The first is attribution: that Anthropic, a developer of large language models (LLMs), has put forward an argument for controlling the pace of AI development — a stance that runs against the grain of an industry defined by racing release schedules. The second is projection: that China could take part in such a slowdown. The second part matters because a moderation of pace pursued by some developers alone would carry little practical force if other major players continued at full speed.

What the Sources Show

The primary sources support the first component: the speed-control argument is genuinely Anthropic's, and the attribution holds. On the second component, the sources treat Chinese participation as an open possibility — a scenario that has been raised and weighed — but they stop short of establishing any commitment by China to a coordinated slowdown of its own AI development.

The Caveat

The gap between the two components explains the qualified rating. The attribution checks out; the forecast cannot be verified beyond its own terms. Whether China joins any effort to moderate the pace of AI development rests on decisions and intentions that have not been made public. A possibility, however plausible, is not a commitment — and a claim that blends an established argument with a contingent outlook cannot be rated as fully true.

Bottom Line

The claim correctly attributes to Anthropic an argument for controlling the speed of AI development, and correctly frames China's participation as a possibility rather than a certainty. On that basis, this desk's verdict is: Mostly True.