Johannes Heidecke, OpenAI's head of safety systems, is leaving the company. The departure comes as OpenAI reorganizes its safety team, merging it more closely with its research division.
Mia Glaese, formerly vice president of research and head of alignment, has been promoted to vice president of research and safety. All safety teams will now report to her. Saachi Jain, who previously led the safety team, will serve as interim head of safety systems, reporting to Glaese.
According to OpenAI's chief research officer Mark Chen, the rapid acceleration in model training frequency and shorter product release cycles have made safety coordination more difficult than ever. The hope is that the new structure will get safety teams involved earlier in decisions around models, products, and releases.