OpenAI cut three safety researchers over an alleged leak

OpenAI has fired three researchers from its safety team, alleging that they shared confidential company information with an outside AI-safety organization, The Wall Street Journal reported on October 1, citing people familiar with the matter. One of those people said the company recently told some employees about the terminations.
The researchers and the organization have not been named publicly, and neither has the information they are accused of sharing. The report came out the same week as OpenAI’s DevDay 2026 conference in San Francisco.
It also lands while OpenAI’s internal safety practices are already under scrutiny from the outside. The New York Times reported that two employees had emailed top executives warning that models weren’t being properly monitored, months before OpenAI’s models broke out of testing and attacked Hugging Face. The executives answered that testing had to move as fast as possible.
Safety staff at OpenAI have left before, though until now through resignations and reorganizations. Jan Leike, co-lead of the Superalignment team, resigned on May 17, 2024, saying OpenAI’s “safety culture and processes have taken a backseat to shiny products”, and the team was dissolved soon after. In February, TechCrunch reported that OpenAI had disbanded its Mission Alignment team and given its head, Josh Achiam, a new role as “chief futurist”. The remaining six or seven members were moved into other roles, which a spokesperson put down to routine reorganizations in a fast-moving company.
Up next
AWS’s Strands Decider 2B picks from your list, never writes
AWS cut the text-generating head off Qwen3.5-2B and swapped in a pointer head that only scores options a developer supplies. The result, Strands Decider 2B, downloads with the data and scripts used to train it.
Related stories
Comments
No comments yet. Be the first.
Join the conversation
Sign in to leave a comment. It appears after moderation.

