OpenAI Head of Safety Johannes Heidecke Departing
Original: OpenAI’s Head of Safety Is Leaving the Company
Why This Matters
Multiple safety leadership exits signal structural shifts in AI safety oversight at a critical moment for frontier model development.
OpenAI's head of safety systems, Johannes Heidecke, informed staff this week that he is leaving the company. His exit follows a reorganization merging safety and research teams, with Mia Glaese taking expanded VP of Research and Safety responsibilities.
OpenAI's head of safety systems Johannes Heidecke has told staff he is leaving the company, WIRED has learned. His departure coincides with a structural reorganization announced by chief research officer Mark Chen, in which OpenAI's safety teams will now report to Mia Glaese, who expands her title to VP of Research and Safety. Saachi Jain will serve as interim head of safety systems. In a staff memo, Chen stated: 'The demands on safety continue to increase—we are training models at a much faster cadence, and release cycles have come down greatly in turn.' Heidecke joined OpenAI in 2021 as an AI safety analyst and became head of safety systems in 2024, succeeding Lilian Weng. His exit is part of a broader wave of departures: chief futurist Joshua Achiam, a nine-year safety researcher, also announced his departure this week. CEO of AGI deployment Fidji Simo also stepped down following medical leave, with Greg Brockman absorbing her responsibilities. These changes come as OpenAI launched GPT-5.6, its most capable agentic coding model, which the company acknowledged exhibited 'concerning forms of misaligned behavior' compared to prior models.