OpenAI Fired Three Safety Researchers, Citing Access to an Executive's Email
austinc3301 · x · 2026-10-09
Former OpenAI safety researcher jasminewang publicly said she and two safety colleagues were fired last week, with the only stated reason being that she accessed an executive's email — part of a pattern, she argues, of OpenAI's "smoke and mirrors" when people disappear. Prominent alignment researcher TurnTrout shared the post, writing that OpenAI "says they're committed to safety but they break the commitments over and over and over." The episode adds to a long-running narrative of safety team turmoil at OpenAI.
Related event: OpenAI Fires Safety Researchers, Sparking Debate Over Internal Checks(15 posts)→
More from Companies & People
- Anthropic Academy launches 6 free AI courses, pitched as a $10,000-degree replacement — Div_pradeep · 2026-10-09
- 'Did not age well': old 'I'm allowed to criticize OpenAI' post resurfaces as drama — dhadfieldmenell · 2026-10-09
- Alex Karp on Palantir's talent philosophy: 17 years mocked, but 'they have the best people' — eliano · 2026-10-09
- Claim that "Gemini is just an agent harness adding Claude models" called out as misleading — prajdabre · 2026-10-09
- xAI's Android App Undergoing Full Rewrite, Reply Bug Fix Ships Same Day — AndrewCurran_ · 2026-10-09
- AI writer yourgirlhils joins Anthropic to lead its AI Fluency effort — round · 2026-10-09