OpenAI's head of safety systems, Johannes Heidecke, is reportedly leaving the company as part of a wider restructuring of its safety and research teams, according to a report by Wired. Heidecke, who joined OpenAI in 2021, informed staff of his departure in an internal memo. The move matters because it signals a shift in how the company organises its safety work, folding it more closely into the development of its frontier AI models.

Under the reorganisation, Saachi Jain, who has previously led OpenAI's safety teams, will step in as interim head of safety systems, while the safety teams will report to Mia Glaese, the newly appointed vice-president of research and safety. Chief research officer Mark Chen said it was important that safety work be "integrated with frontier-model development", playing an earlier and more direct role in shaping model, product and launch decisions. The changes follow the recent release of OpenAI's latest model, GPT-5.6, which was approved by the US government, and the company also retains a Head of Preparedness hired earlier this year to mitigate severe risks.