🛡 OpenAI Head of Safety Leaves Company
OpenAI has undergone a major leadership reorganization. Johannes Heidecke is leaving his post as Head of Safety Systems, and the safety teams will now report to Mia Glaese. Safety will become an integral part of the model development process.
🌍 A shift in focus from isolated safety auditing to the deep integration of safety protocols into the development cycle of frontier models. While this may accelerate releases, it also increases risks.
👤 The change in management structure could impact the pace and transparency of releasing new models, such as GPT-5.6.
Source 1: https://www.wired.com/story/openai-head-of-safety-leaving/