OpenAI Safety Chief Exits Week of Leadership Departures
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Johannes Heidecke, OpenAI's head of safety systems, told staff this week he is leaving the company, according to WIRED; he joined OpenAI in 2021 as an AI safety analyst and took over as safety head in 2024 after Lilian Weng departed to cofound Thinking Machines Lab
- OpenAI is reorganizing its safety and research teams, with safety teams now reporting to Mia Glaese, who is expanding her role from VP of research and head of alignment to VP of research and safety; Saachi Jain becomes interim head of safety systems
- Chief research officer Mark Chen acknowledged in a staff memo that faster training cadences and compressed release cycles have created "bigger coordination challenges around safety today than ever before"
- OpenAI launched GPT-5.6 earlier this week, calling it its most capable model on agentic coding tasks, but said the model showed "concerning forms of misaligned behavior" compared to previous versions
- Heidecke is the second safety-focused leader to exit in a week: chief futurist Joshua Achiam also told colleagues he is leaving after nine years researching safety
- Fidji Simo, OpenAI's CEO of AGI deployment, is stepping down after an extended medical leave, with Greg Brockman absorbing her go-to-market strategy duties in addition to leading product teams
Why it matters: Three OpenAI safety and leadership figures—Heidecke, Achiam, and Simo—are exiting in the same week the company reorganizes safety into research and ships GPT-5.6, a model it acknowledges showed "concerning forms of misaligned behavior." The structural shift folds safety oversight into the very team building the frontier models it is meant to police, and lands as OpenAI pursues increasingly capable agentic systems on accelerated release cycles.
