Anthropic Admits Submitting False, Outdated Info to Congressional Inquiry on AI Risk
JeffLadish · x · 2026-09-06
- Ethan Perez confirmed Anthropic submitted factually false, outdated conclusions to a Congressman's public inquiry, and has since posted a corrected account of its current understanding.
- He promised (1) a more detailed alignment assessment of the incidents and (2) a direct follow-up to the letter.
- Critics note that even if accidental, Anthropic's policy team may have spread misinformation downplaying AI risks, and want the internal process failure explained.
More from Companies & People
- ScienceClaw Hackathon at MIT Media Lab to build decentralized AI agent collectives for science — ProfBuehlerMIT · 2026-10-23
- AI-generated game clones spook indie developers, who now hesitate to show their work — emmanuelvivier · 2026-10-23
- Lenovo adds 5 AI Library use cases, claims 1-week deployment and up to 70% lower token cost — shashib · 2026-10-23
- CuspAI hires Application Scientists for AI-driven materials discovery in energy and electronics — jonfildes · 2026-10-23
- ELLIS Alicante opens 2-year postdoc position on Human-AI Teaming and Agentic AI — HildeKuehne · 2026-10-23
- Ex-Googler shares contract: Google UK non-disparagement clause bars criticism forever — PMinervini · 2026-10-23