OpenAI fired three researchers fighting to keep chain-of-thought visible after requesting their Signal chats
JeffLadish · x · 2026-10-23
Reporting indicates the three researchers OpenAI fired last month had spent the summer fighting to keep AI chain-of-thought visible; OpenAI then asked to see their Signal chats, and a week later they were let go.
Safety researcher Jeff Ladish draws structural lessons:
- Safety-critical industries like nuclear and finance give employees defined confidential channels to raise concerns with auditors and regulators; AI needs similarly clear rules governing labs, outside evaluators and employees
- Calling an evaluator an 'independent auditor' means little unless its access and employees' ability to speak candidly are protected
- OpenAI says it supports embedded auditing yet fired employees for talking to those auditors — showing the weakness of leaving protections to voluntary, lab-controlled arrangements
- Employees never clearly told which channels are permitted can't comply, and everyone watching learns that talking to evaluators is risky
Related event: OpenAI Fired Three Safety Researchers After Requesting Their Signal Chats(5 posts)→
More from Companies & People
- Anima Anandkumar named to AI100 list, featured on Nasdaq tower in Times Square — AnimaAnandkumar · 2026-10-23
- 'Succession' Writer's AI-Titan Satire The Audacity Arrives on AMC — nordicinst · 2026-10-23
- Rambus adopts ChipAgents companywide and invests to advance AI-driven chip design — WilliamWangNLP · 2026-10-23
- UCLA roommates-turned-founders: the story behind sidewalk delivery robot startup Coco Robotics — zhoubolei · 2026-10-23
- MiniMax partners with SandBase on city-remix hackathon with $3,000 grand prize — MiniMax_AI · 2026-10-23
- AI Lab Matrix: Stanford Hazy's 13 people founded 7 startups, Berkeley BAIR 23 total — FinanceYF5 · 2026-10-23