Jasmine Wang, Tomek Korbak, and Mikita Balesni, three safety researchers OpenAI fired last week, have published an open letter denying the company's misconduct allegations and warning that their dismissals are creating a chilling effect on internal safety work.
OpenAI terminated the three after an investigation found a "pattern of misconduct" involving mishandling of research information in violation of company policy. The company alleged they accessed and handled sensitive information outside established procedures. OpenAI did not specify which policies were violated or provide details about the dismissals when asked by TechCrunch.
The researchers deny wrongdoing and argue their firing signals a dangerous cultural shift at OpenAI. They write that safety work at an AI company requires "close collaboration with outside experts" and that employees now fear speaking up despite behavior that was normal a month ago suddenly becoming grounds for termination.
Wang explained on X that she was fired for accessing an executive's email that OpenAI had delegated to her for recruiting purposes. When she no longer needed access, she requested IT remove it. The request was not processed. She says she accidentally opened a sensitive email, reported it to the executive within minutes, and asked IT again to remove the access. She characterizes the stated reasons for termination as "not adding up."
Korbak's actions related to the Hugging Face incident, in which agents breached external systems. The researchers argue internal policies were being developed in real time during that investigation, and Korbak believed he was acting within company norms by communicating with outside safety evaluators.
Balesni coordinated with board members and executives on an AI monitorability problem, according to the letter. He removed sensitive details before sharing materials and checked in with his reporting line.
OpenAI declined to directly address questions about specific policy violations or how it protects employees who raise safety concerns. An internal memo attributed to a research leader praised the three researchers' contributions and denied the firings were retaliation for raising safety concerns. An OpenAI spokesperson told TechCrunch the pattern of misconduct extends beyond sharing information with an outside AI evaluation group but provided no further specifics.
The three researchers called on OpenAI to embed third-party safety auditors within the organization, preserve monitorability of frontier models, and maintain an open culture between safety researchers and the broader safety ecosystem. OpenAI's memo states it agrees with those recommendations.