OpenAI’s three fired safety researchers, Jasmine Wang, Tomek Korbak, and Mikita Balesni, argued in an open letter that they didn’t break company rules on sensitive information, the reason OpenAI gave last week for firing them amid allegations that included passing material to an outside AI safety group. The three wrote that the abrupt firings have left colleagues afraid to raise safety concerns or work with outside evaluators. Wang separately posted on X that OpenAI cited her access to an executive’s inbox, which she said it had granted her for recruiting and failed to revoke when she asked. OpenAI shared a memo with TechCrunch denying the firings were retaliation, and a spokesperson said its investigation found a “pattern of misconduct” beyond the outside sharing.
