EN ▾
SNAPE REP RT Front page →
ConfirmedAI safety18 Nov, 15:04 UTC

Pinker, Tegmark and 'Beff Jezos' sign statement calling Anthropic 'not a safe AI company'

File image of the Anthropic logo
File image via Wikimedia Commons

The Center for AI Safety's MisAnthropic Statement, issued ahead of the IPO, says Anthropic races to build self-improving AI while weakening its safety pledges.

Read at The MisAnthropic Statement (Center for AI Safety) Share

Why it matters

The statement lands a day before Anthropic's expected Nasdaq debut and unusually unites safety researchers with accelerationists such as Guillaume Verdon and George Hotz. Its evidence section says Anthropic weakened its Responsible Scaling Policy in February, dropping unilateral pause commitments, and lobbied against SB 1047, the RAISE Act and the AEGIS Act. Dan Hendrycks says other safety groups agreed privately but would not sign for fear of losing funding. Watch whether Anthropic responds before pricing and whether IPO investors raise it.

Key facts

  • Organised by the Center for AI Safety
  • Published at misanthropicstatement.com on 18 Nov
  • Signers include Steven Pinker, Max Tegmark, Judea Pearl and Thomas Wolf
  • Also Guillaume Verdon, George Hotz and Igor Babuschkin
  • Says Anthropic weakened its Responsible Scaling Policy in February
  • Cites lobbying against SB 1047, the RAISE Act and the AEGIS Act

Sources

More on AI safety
More from Snape ReportFront page →