OpenAI fired three safety researchers over Signal disappearing messages amid leak probe
xuanalogue · x · 2026-10-23
Per WSJ reporting, OpenAI investigators concluded Ryan Greenblatt was the source for The Information's story on 'recurrent depth' — a technique boosting capabilities at the cost of monitorability. OpenAI then fired three safety researchers after finding disappearing messages enabled on their personal-phone Signal chats with Greenblatt.
- The three had spent the summer fighting to keep chain-of-thought visible; OpenAI asked to see their Signal chats, and a week later they were summoned to simultaneous 2 p.m. meetings and let go
- Fired researcher Tomek Korbak said he was told it was a 'breach of trust' via unsanctioned channels
- Greenblatt denies leaking: 'I take confidentiality very seriously'
- Garrison Lovely argues there are reasons to believe him: safety researchers are the most skittish about leaking, and third-party researchers' access depends entirely on the company
Related event: OpenAI Fired Three Safety Researchers Over Vanishing Signal Chats(8 posts)→
More from Companies & People
- Senior researcher rebuts 'one-person pretraining' fear: your fundamentals matter more, not less — RexDouglass · 2026-10-23
- Anthropic proactively issues batch refunds and credit notes to a user — TechPreacher · 2026-10-23
- Manycore Aholo Global 3D AI Hackathon opens with $47,800 prize pool — Scobleizer · 2026-10-23
- Thousands built claw machines across Europe as the tour wrapped at GTC Berlin — lifebypixels · 2026-10-23
- Instinct personal agent growing 10% daily, yet resort board had never heard of it — johncoogan · 2026-10-23
- WSJ: OpenAI fired 3 safety researchers over disappearing Signal chats on model Astra — rohanpaul_ai · 2026-10-23