Latest
AI Affairs, home

Thursday 19 November 2026

Markets

Coalition challenges Anthropic’s safety branding ahead of planned IPO

The Center for AI Safety’s public statement accuses Anthropic of weakening safety commitments and seeking rules that favour its development plans.

Dan Hendrycks and another speaker seated onstage at 2025 TechCrunch Disrupt
Photo: TechCrunch, CC BY 4.0, via Wikimedia Commons (cropped)

The Center for AI Safety challenged Anthropic’s safety branding on 18 November ahead of its planned IPO, as its executive director said the company was asking the public for “tens of billions of dollars”, according to a Business Wire release issued in San Francisco. The challenge takes the form of the MisAnthropic Statement, a public declaration signed by researchers, legal scholars and technology executives.

Key points

  • The coalition accuses Anthropic of weakening safety pledges as it pursues AI capable of improving its own development.
  • Dan Hendrycks, the Center for AI Safety’s executive director, linked the criticism to what he described as a request for tens of billions of dollars from the public.
  • The signatories include AI safety researchers, open-source advocates, ethicists and accelerationists.

MisAnthropic Statement challenges Anthropic’s investor pitch

The statement accuses Anthropic of presenting itself as a safety-focused company while seeking to build recursively self-improving AI — systems capable of improving their own AI development. It says Anthropic warns that such systems could concentrate power and escape human control, yet is racing to develop them. Business Wire reproduced the statement’s charge that Anthropic had “quietly weakened its safety pledges and lobbied for self-serving regulation”.

The coalition’s objection reaches Anthropic as it prepares to seek public-market investors. Hendrycks said Anthropic could not credibly claim to be safe after weakening earlier pledges as it asked the public for tens of billions of dollars to develop recursively self-improving AI.

The statement also alleges that Anthropic uses its safety reputation to attract investors, employees and the public. It objects to the company’s attempt to present itself to policymakers as a responsible developer while pursuing a technology that the coalition considers dangerous.

Responsible Scaling Policy and legislative allegations

The statement says Anthropic weakened its Responsible Scaling Policy in February, including by abandoning unilateral commitments to pause development, LavX reported. It also alleges that Anthropic promoted recursive self-improvement to prospective public-market investors as a competitive advantage, despite warning about the risks of the same technology.

The coalition further accuses Anthropic of lobbying to weaken SB 1047, the RAISE Act and the AEGIS Act. According to the statement, the company also pressed in private meetings for federal preemption of state AI laws, which would replace state-level requirements with federal rules.

Hendrycks said the people backing the statement disagreed on many other questions. Their shared objection, he said, was to a company treating its own dominance as a public good. The statement asks readers to reject Anthropic’s approach to building recursively self-improving AI.

CAIS signatories span opposing AI camps

The signatories come from groups that often take different positions on AI development. Business Wire identified safety researchers, open-source advocates, AI ethicists and accelerationists among them. The list includes Steven Pinker and Nobel economics laureate Oliver Hart of Harvard University, Turing Award laureate Judea Pearl of UCLA, and Harvard Law scholars Laurence Tribe and Lawrence Lessig.

Technology signatories include Thomas Wolf, co-founder and chief science officer of Hugging Face, and Guillaume Verdon, founder and chief executive of Extropic and founder of the e/acc movement. The release also names Rumman Chowdhury, founder and chief executive of Humane Intelligence, Tristan Harris, co-founder and president of the Center for Humane Technology, and George Hotz, founder and chief executive of the tiny corp.

Other signatories named in the release include MIT physics professor Max Tegmark and Igor Babuschkin, co-founder and chief executive of River AI. Mathematicians Peter Scholze, James Maynard and David Mumford, each a Fields Medalist, also signed the statement.

Sources

Topics: Foundation models, IPO, Safety

What should we cover? Tell us

The weekly briefing

The AI Affairs weekly briefing: the week's 10 most important AI stories in business, government and markets, every Monday. We'll email you a link to confirm. Every email has a one-click unsubscribe. Privacy policy