OpenAI's head of security system resigns, adding another member to the wave of senior departures.

According to reports, Johannes Heidecke, the head of security at OpenAI, informed colleagues this week that he will be leaving the leading artificial intelligence (AI) company. This is the latest in a series of recent departures of key executives at OpenAI, which is part of a restructuring of the security team to cope with the increasingly rapid pace of model iteration. Citing an internal memo, it was announced that Mark Chen, Chief Research Officer at OpenAI, has declared that the security team will now report to Mia Glaese, Vice President of Research and Alignment. Glaese's role has been significantly expanded in this reorganization, as she will now serve as the Vice President of Research and Security, overseeing both research and security departments. Meanwhile, Saachi Jain, who previously led the security team, will serve as the acting head of security, reporting to Glaese. Chen explained the reasons for the restructuring in the memo, stating that security requirements are increasing, model training pace is accelerating, and release cycles are shortening, making the coordination challenges around security issues more severe than ever before. He stated in a statement, "We thank Johannes for his contributions to OpenAI. Integrating security work with cutting-edge model development is essential to ensure that security plays a more direct role in critical model, product, and release decisions. We look forward to Mia Glaese leading the new chapter in research and security." Heidecke joined OpenAI in 2021 as an AI security analyst, and took over as head of security after the departure of Lilian Weng in 2024 to start Thinking Machines Lab. His departure comes as OpenAI is striving to release increasingly powerful models. Just this week, the company released its most powerful model, GPT-5.6, for intelligent agent programming tasks, but compared to previous models, the official acknowledgment that GPT-5.6 exhibits concerning alignment bias issues. Heidecke's departure is not an isolated incident, as there have been a series of changes in the research and product leadership at OpenAI. Earlier this week, Joshua Achiam, the chief futurist at the company, who has been involved in security research for nine years, also informed colleagues of his departure. In addition, Fidji Simo, Chief Executive Officer responsible for AGI deployment at OpenAI, announced this week that he will be stepping down after an extended period of illness. The company stated that Greg Brockman will continue to lead the product team and will also take on the market entry strategy responsibilities.