OpenAI Fires Three Safety Researchers After Internal Investigation; Researchers Dispute Reasons

OpenAI has fired three researchers who worked on AI safety: Jasmine Wang, Tomek Korbak and Mikita Balesni. The company said the dismissals followed an internal investigation into alleged violations of its policies on sensitive information.

The facts behind the investigation have not been made public, so the competing claims cannot be independently verified.

The case adds to a long-running debate about safety culture at leading AI labs:

AI companies must protect confidential information, such as unreleased model details and security-sensitive research. They also face pressure to show they welcome internal criticism about risks. How companies handle disputes like this affects trust among employees, regulators and the public.

The dismissals came in the same week OpenAI launched GPT-6, and as reports questioned how its annualised revenue is calculated. Axios reported about $50 billion rather than a previously circulated $70 billion, a difference attributed to how revenue from cloud partners is counted.

Watch for any further statements from OpenAI or the researchers, possible legal action, and whether regulators or lawmakers cite the case in debates on AI whistleblower protections.