Anthropic's risk statement slammed for self-praise over frank talk on AI risks

ShakeelHashim · x · 2026-09-10

Anthropic issued a statement about "the last 24 hours," stressing its transparency, industry-leading safeguards, pioneering work in mechanistic interpretability, and its Responsible Scaling Policy.

Sharing the statement via Hadas Gold, Shakeel Hashim called it "a very bad statement": instead of speaking frankly about risks or reiterating what Hubinger and Coxon said, Anthropic spent much of it patting itself on the back — a depressing level of corporate comms from a company that has traditionally been above that.

Related event: Anthropic's Crisis Statement Slammed as Self-Congratulatory PR(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →