Why OpenAI keeps losing its top safety leaders one after another

Johannes Heidecke, OpenAI’s head of safety systems, has told staff he is leaving the company, according to a report by Wired. His exit comes immediately after an internal reorganisation designed to merge OpenAI’s safety and research divisions, and it is the latest in a pattern of safety-focused leaders departing the company.

Heidecke’s five-year stint at OpenAI, which began in 2021, ends two years after he stepped into the safety chief role in 2024, replacing Lilian Weng, who left to co-found rival startup Thinking Machines Lab. Just days before Heidecke’s announcement, OpenAI’s Chief Futurist, Joshua Achiam, also revealed he was leaving after nine years of safety research.

In an internal memo, Chief Research Officer Mark Chen said the company’s safety teams would now be consolidated under Mia Glaese, who takes on an expanded role as VP of Research and Safety, with Saachi Jain stepping in as interim head of safety systems. ‘We have bigger coordination challenges around safety today than ever before,’ Chen wrote, citing a faster training cadence and shorter release cycles.

The departures aren’t limited to the safety division. Fidji Simo, CEO of AGI Deployment, is also stepping down after an extended medical leave, with co-founder Greg Brockman absorbing her responsibilities. The shakeup reflects the mounting operational pressure inside OpenAI as it races to deploy advanced AI models while eyeing an IPO.

Image: Wikimedia Commons/by Loic Le Meur