OpenAI head of safety Johannes Heidecke is leaving, Wired reports

OpenAI’s head of safety, Johannes Heidecke, is leaving the company, Wired reports today. The move lands as OpenAI seeks to further integrate its research and safety teams, a shift Wired says could reshape how risk is handled around rapidly evolving models. It also comes amid intensifying scrutiny of AI governance from regulators and competitors as frontier systems roll out at pace.

Wired reports that Heidecke helped steer safety efforts during a period of fast product launches and internal reorganization, and that OpenAI has not yet named a successor while emphasizing cross-functional safety work as it scales and aligns safety with core product engineering. Watch for OpenAI to identify interim leadership and detail how safety review will be embedded across engineering and research in the near term. [Developing. This article will be updated.]