Mother Jones: OpenAI's 'Independent Assessment' Doesn't Look Independent at All
motherjonesmag · reddit · 2026-11-18
Mother Jones reports that OpenAI's "Independent Assessment" of its AI systems' safety and oversight doesn't look independent at all.
The article argues the assessment's conclusions on oversight, control, and security were produced under arrangements too close to OpenAI itself to count as objective external review — part of a long-running criticism of the company's self-supervised governance, where "independent" framing masks conflicted evaluation of its own safety claims.
More from Companies & People
- SpaceX CFO praises Grok Bot as 'probably the best agentic AI solution out there' — elonmusk · 2026-11-18
- Factory joins US Commerce-approved Secure AI by Design Consortium with NVIDIA, OpenAI, IBM — annbordetsky · 2026-11-18
- Firecrawl opens first office outside SF, in Toronto, and is hiring — devdigest · 2026-11-18
- Forecaster chatter claims Jensen Huang resigned on the 12th — unverified — NathanpmYoung · 2026-11-18
- "If Chinese AI collapsed without Jensen, it never needed him" — teortaxesTex fires back — teortaxesTex · 2026-11-18
- Greg Brockman enters "full management mode" at OpenAI as IPO nears — steph_palazzolo · 2026-11-18