3 ex-OpenAI researchers say safety concerns led to their firing
What's new: The dispute lands as OpenAI faces scrutiny over agent behavior and plans to expand third-party safety reviews.
Three former OpenAI researchers say the company fired them after they raised concerns about AI safety, a claim OpenAI rejects. Mikita Balesni, Tomek Korbak and Jasmine Wang, who worked on safety or alignment teams, published an open letter saying the dismissals have left employees afraid to speak openly. OpenAI said an internal investigation found they mishandled sensitive information outside company procedures and called the case a major breach of trust. The clash comes as OpenAI faces wider scrutiny over how it develops autonomous AI agents and whether it will follow through on a Sept. 12 pledge by Sam Altman to broaden access for third-party safety evaluators.