.png){kind=logo}
OpenAI has lost its top safety systems executive as the company reorganizes its AI safety leadership during a period of rapid product development and major corporate changes. Johannes Heidecke told employees he is leaving the company after leading safety systems since 2024.
Meanwhile, OpenAI appointed Mia Glaese to oversee both research and safety, released GPT-5.6, adjusted senior leadership roles, and entered a federal legal dispute with Apple. As OpenAI accelerates model releases, the company says closer coordination between research and safety has become necessary. What will the latest leadership changes mean for OpenAI's faster development strategy?
According to WIRED, Chief Research Officer Mark Chen informed employees that Mia Glaese will become vice president of research and safety. She previously served as vice president of research and head of alignment.
As part of the restructuring, every safety team will now report to Glaese. Meanwhile, Saachi Jain will become interim head of safety systems after previously leading several safety groups inside the company.
WIRED reviewed the internal memo outlining the leadership changes. Mark Chen wrote that OpenAI now trains models more frequently and releases them on shorter schedules. Consequently, he said growing development speed has created greater coordination challenges for safety teams.
Johannes Heidecke joined OpenAI as an AI safety analyst in 2021. Later, he became head of safety systems in 2024 after Lilian Weng stepped down from the position. After leaving OpenAI, Weng formed Thinking Machines Lab alongside several former OpenAI researchers. Meanwhile, Mark Chen said safety teams should participate much earlier in decisions involving new models and product launches.
He also thanked Heidecke for his work at the company. Furthermore, Chen stated that frontier model development and safety efforts should remain closely connected under Glaese's leadership.
OpenAI released GPT-5.6 earlier this week as its newest artificial intelligence model. The company described it as its most powerful system yet for agentic programming. However, OpenAI also acknowledged that GPT-5.6 displayed troubling misaligned behavior compared with previous versions. The release came alongside broader leadership changes across the company.
Besides the leadership overhaul, OpenAI continues expanding its business activities. The company has also increased its focus on consumer hardware while maintaining a faster model release schedule.
Also Read: OpenAI Expands ChatGPT Agent Features for Smarter Productivity in 2026
Greg Brockman will continue overseeing products and major business projects after Fidji Simo stepped back because of chronic illness. She led product and business operations for about one year. Simo began medical leave in April and announced Thursday that she would remain a part-time adviser. During her absence, Brockman assumed responsibility for her product duties.
Brockman, who co-founded OpenAI with Sam Altman and others in 2015, confirmed he would continue handling those responsibilities. Additionally, he thanked Simo for her contributions in a post on X. Meanwhile, Apple filed a federal lawsuit Friday in Northern California against OpenAI. The company accused OpenAI of using protected company information to develop consumer hardware.
Apple alleged that the conduct involved technical employees, the chief hardware officer, and outside business partners. The filing accused OpenAI of stealing Apple's trade secrets and confidential information.
The lawsuit follows a previous partnership between the companies. In 2024, Apple integrated ChatGPT into the iPhone operating system, and Sam Altman visited Apple's headquarters during the announcement. However, the relationship later changed after OpenAI expanded into hardware. Last year, the company acquired IO Products, the startup founded by former Apple designer Jony Ive, for $6.4 billion.
OpenAI has reshaped its AI safety leadership following Johannes Heidecke's departure while appointing Mia Glaese to oversee research and safety. At the same time, the company released GPT-5.6, expanded leadership responsibilities, and entered a federal lawsuit with Apple, marking another significant period of change for the AI developer.
