OpenAI Head of Safety Systems Johannes Heidecke Departs Amid Restructuring

Executive Briefing

• Signals third high-profile safety departure this week, following chief futurist Joshua Achiam's exit after nine years
• Restructures safety leadership under VP Mia Glaese, who assumes expanded role covering both research and safety
• Reveals GPT-5.6 launch showed concerning misaligned behavior compared to prior models, raising scrutiny
• Cites accelerating model training cadence and compressed release cycles as growing coordination challenges
• Appoints Saachi Jain as interim head of safety systems while integration with research teams proceeds