Ex-OpenAI employee says external safety assessors got far less access than Altman promised
j_asminewang · x · 2026-11-17
A departing OpenAI employee claims that in their final week, external auditors were still negotiating hard for minor improvements to access and scope of safety-relevant information — despite Sam Altman publicly committing to "employee-like access" for independent assessors.
OpenAI simultaneously announced embedded assessments with METR, Apollo Research, Transluce and Goodfire, but the whistleblower notes the terms fell short:
- No Slack access, where all of OpenAI's safety debates happen
- OpenAI itself picks which employees assessors may talk to
The author suggests employees failed to hold leadership to Altman's public commitment, attaching a screenshot of his original pledge as evidence.
More from Companies & People
- Stanford HAI awards $3.5M to seven teams via 2026 Hoffman-Yee Research Grants — StanfordHAI · 2026-11-17
- Runway CTO Kamil Sindi to break down training and real-time serving of video foundation models at AI Dev NYC — DeepLearningAI · 2026-11-17
- Iliad January Fellowship Applications Close Today: Paid Math-Based AI Alignment Research — austinc3301 · 2026-11-17
- Kai-Fu Lee: CEOs should use AI agents as Chief of Staff or even Co-CEO — kaifulee · 2026-11-17
- Omar Khattab on impactful AI research: invest in projects, not papers — lateinteraction · 2026-11-17
- ARK Newsletter Flags Jensen Huang's Resignation as Major Test for Nvidia — downingARK · 2026-11-17