Former OpenAI safety researchers Mikita Balesni, Tomek Korbak, and Jasmine Wang stated they were terminated last week for prioritizing safety over short-term corporate interests. In an open letter to company leadership, the trio denied allegations of leaking company secrets and urged OpenAI to honor commitments to third-party safety audits and frontier model monitorability. OpenAI previously attributed the dismissals to an internal investigation finding that the researchers handled sensitive information outside established processes. The fired employees contest this characterization, framing their termination as retaliation for advocating stricter safety protocols within the organization.