OpenAI Fires Three Safety Researchers, Raising Concerns About AI Collaboration and Risk Assessment

OpenAI has parted ways with three safety researchers, sparking a major dispute over the company's commitment to independent AI safety oversight and external collaboration. Tomek Korbak, Jasmine Wang, and Mikita Balesni were terminated in early October 2026 after an investigation allegedly found they violated company policies on handling sensitive information. However, the researchers have denied wrongdoing and published an open letter claiming they were terminated for their work collaborating with third-party safety evaluators, signaling a troubling retreat from safety-focused partnerships.