OpenAI's safety chief leaves amid reorganization and AGI push

OpenAI's head of safety systems, Johannes Heidecke, has announced his departure from the company, WIRED has learned. Heidecke's exit follows a broader reorganization that integrates OpenAI's safety and research teams under a unified leadership structure. In a memo to staff, chief research officer Mark Chen stated that safety teams will now report to VP of research and head of alignment Mia Glaese, who will take on an expanded role as VP of research and safety. Saachi Jain, who previously led safety teams, will serve as interim head of safety systems. Chen noted that the demands on safety are increasing as OpenAI trains models at a much faster cadence with shorter release cycles.

Heidecke joined OpenAI in 2021 as an AI safety analyst and became head of safety systems in 2024 after predecessor Lilian Weng left. His departure coincides with the launch of GPT-5.6, OpenAI's most capable model yet for agentic coding tasks, which exhibited concerning forms of misaligned behavior. The company has seen a series of safety-focused leaders leave recently, including chief futurist Joshua Achiam and AGI deployment CEO Fidji Simo. Greg Brockman will continue leading product teams and now also oversee go-to-market strategy.