Pinker, HF co-founder Wolf join rare cross-camp statement slamming Anthropic's safety branding ahead of IPO
wfithian · x · 2026-11-18
Ahead of Anthropic's IPO, a rare coalition spanning AI safety, open source, e/acc, science, and law signed the "MisAnthropic Statement" — signatories include Steven Pinker, Hugging Face co-founder Thomas Wolf, Laurence Tribe, Turing laureate Judea Pearl, two Nobel laureates, and xAI co-founder Igor Babuschkin.
Core claims:
- Anthropic warns recursively self-improving AI could concentrate power and escape control, yet races to build it first.
- The company uses its safety branding to court investors, employees, and the public while quietly weakening safety pledges and lobbying for self-serving regulation.
- Conclusion: Anthropic is not a safe AI company, and its power grab in the name of the greater good must be rejected.
Signatory wfithian says he has three young children and sees little evidence Anthropic is seriously coordinating a stop to activity its own leaders deem dangerous.
Related event: Pinker, Thomas Wolf Sign Statement Blasting Anthropic's Safety Rhetoric(2 posts)→
More from Companies & People
- NOETIK partners with Memorial Sloan Kettering to build cancer biology world models — anshulkundaje · 2026-11-18
- Audit is the economy's evaluation layer; signature and liability become the product — krishnan · 2026-11-18
- KPMG names AI audit limits: can't run thousands of audits, client data too dirty — krishnan · 2026-11-18
- Where KPMG's audit agents broke: mapping regulations to evidence still needs humans — krishnan · 2026-11-18
- KPMG's audit agents: built in 4 months, handled data, analysis and drafting — krishnan · 2026-11-18
- Boltz, the AI lab that moved from MIT to the UK, raises $110M — GabriCorso · 2026-11-18