Three safety researchers fired by OpenAI have disputed the company's claim that they mishandled confidential information, saying their dismissals are chilling open safety work at the firm. In an open letter, Jasmine Wang, Tomek Korbak, and Mikita Balesni denied leaking details about OpenAI's newest models and said they acted in good faith under policies that were still being developed.
Wang separately said OpenAI told her she was fired for accessing an executive's email, which she said was delegated to her for recruiting and accidentally opened. The researchers argue that the abrupt terminations have made employees afraid to raise concerns or collaborate with outside safety groups, and they called on OpenAI to keep third-party auditors and preserve monitorability of frontier models.
OpenAI has not formally responded to the letter, but an internal memo attributed to a research leader praised the three researchers' contributions and said the firings were not retaliation for speaking out. A spokesperson, however, said an investigation found a 'pattern of misconduct' in clear violation of policies on mishandling research information. OpenAI did not specify which policies were violated.