OpenAI fired three safety researchers for allegedly sharing confidential information with an outside AI safety organization, violating company policies. The dismissals follow months of security incidents involving rogue AI agents, safety team departures, and ongoing concerns about the company's commitment to AI safety culture.