A public statement titled “The Mis Anthropic Statement” was released recently, challenging an apparent conflict between Anthropic’s safety commitments and its commercial expansion. According to Odaily, the statement accuses Anthropic of racing to develop recursively self-improving AI and pursue superintelligence while presenting itself as an AI safety company and using that reputation to advance commercial and political interests.

The statement says Anthropic has warned that recursively self-improving AI could concentrate power and escape human control, yet continues to advance the technology. It also alleges the company uses its safety brand to attract investors, employees and the public while weakening its safety commitments and lobbying for regulation favorable to its interests. The signatories argue Anthropic is not a safe AI company and call on the public to reject what they describe as its expansion of power in the name of public interest.

Signatories include Harvard psychology professor Steven Pinker; Hugging Face co-founder Thomas Wolf; Nobel laureate in economics Oliver Hart; Nobel laureate in physics Sheldon Lee Glashow; Turing Award recipient Judea Pearl; xAI co-founder Igor Babuschkin; Center for Humane Technology co-founder Tristan Harris; and Future of Life Institute founder Max Tegmark. The statement was also signed by several Fields Medal recipients. Anthropic is pursuing an initial public offering (IPO).