Hours before Anthropic's shares began trading on the Nasdaq, a coalition of more than 80 scientists, legal scholars and technology figures published a public statement with a blunt message: the company marketing itself as the industry's safety leader is, in their words, "not a safe AI company." The MisAnthropic Statement, hosted at misanthropicstatement.com and announced on November 18 with the backing of the Center for AI Safety, unites signatories who agree on almost nothing else.
The name list is the story. It includes Nobel economics laureate Oliver Hart, Turing Award winner Judea Pearl, Nobel physicist Sheldon Lee Glashow, Harvard psychologist Steven Pinker, constitutional scholars Laurence Tribe and Lawrence Lessig, MIT physicist and Future of Life Institute founder Max Tegmark, three Fields Medalists — Peter Scholze, James Maynard and David Mumford — as well as Hugging Face co-founder Thomas Wolf, xAI co-founder Igor Babuschkin, e/acc movement founder Guillaume Verdon, hacker-entrepreneur George Hotz, Center for Humane Technology co-founder Tristan Harris and AI critic Gary Marcus. Effective accelerationists, safety researchers, ethicists and open-source advocates rarely share a letterhead; CAIS executive director Dan Hendrycks says that is the point.
The core charge is a gap between branding and behavior. The statement argues Anthropic warns that recursively self-improving AI could concentrate power and escape human control, yet is racing to build exactly that capability — and pitching it to public-market investors as a competitive advantage while doing so. It further accuses the company of using its safety reputation to court investors, employees and regulators, and of lobbying for rules that favor its own development plans, including efforts to weaken California's SB 1047, the RAISE Act and the AEGIS Act, and pressing in private meetings for federal preemption of state AI laws.
The most concrete allegation concerns February 2026, when Anthropic revised its Responsible Scaling Policy. TIME reported at the time that the revision dropped the company's 2023 pledge never to train a system unless it could guarantee in advance that safety measures were adequate. Chief scientist Jared Kaplan defended the change, arguing unilateral commitments made little sense "if competitors are blazing ahead." The statement leans on that episode to argue the safety pledges investors are buying into have already been softened once.
Timing was deliberate. The statement landed the day Anthropic priced its upsized IPO at 83 dollars per share — above the marketed 70-to-78-dollar range — in what is set to be the largest offering on record, and as the company prepared to ask public investors for tens of billions of dollars. "You cannot credibly claim to be safe while weakening your own previous safety commitments and asking the public for tens of billions of dollars to build recursively self-improving AI," Hendrycks said in the release. Anthropic did not immediately respond to a request for comment.
Realistically, the letter will not dent a debut that was already heavily oversubscribed. But it draws a durable line: the company that built its brand on being the responsible lab now trades as a public company while some of the most credible names in science and law publicly dispute the brand. For a market that has been pricing Anthropic's safety positioning as an asset, the statement is the first organized challenge to that premium from inside the AI debate itself.
Comments (0)
Log in to join the discussion
Log InNo comments yet