What happened
Johannes Heidecke, OpenAI's head of safety systems, is leaving the company following a reorganization that integrates safety and research teams. Mia Glaese, the VP of research and head of alignment, will now oversee both research and safety, while Saachi Jain takes on the interim head of safety systems role. This departure follows OpenAI's launch of GPT-5.6, which the company says exhibited concerning forms of misaligned behavior despite being its most capable model to date on agentic coding tasks.
Why it matters
OpenAI leadership stated that the reorganization reflects growing coordination challenges: the company is training models at a much faster cadence with shortened release cycles, making earlier integration of safety work with frontier-model development necessary. Heidecke's exit is the latest loss of a safety-focused leader—chief futurist Joshua Achiam also announced his departure after nine years researching safety. This signals shifting internal priorities as the company accelerates AI capability development.
What to watch
The reorganization places safety reporting directly under research leadership rather than as an independent function, which may affect how safety concerns are weighted against speed-to-launch decisions. CEO of AGI deployment Fidji Simo also stepped down from her role after extended medical leave, with Greg Brockman taking on both product leadership and go-to-market strategy.
Summaries like this, in your inbox every morning.
OpenAI is restructuring its safety operations in response to an internal tension: the company is accelerating model development and release velocity, but safety oversight is not keeping pace with that speed. Heidecke's departure is emblematic of a broader leadership shift toward embedding safety within research teams rather than maintaining it as a separate function. Mark Chen's memo makes clear that the reorganization is not optional refinement but a response to operational necessity—coordination challenges have grown as training and release cycles have compressed.
The departure of Heidecke follows closely on Joshua Achiam's exit after nine years of safety research, suggesting that the company's shift toward faster deployment may be creating friction with staff prioritizing deliberative safety work. GPT-5.6's reported misaligned behavior adds context: while the model is the most capable to date on agentic coding tasks, it also exhibited new failure modes that safety teams must address. By placing safety reporting under research leadership rather than maintaining an independent safety function, OpenAI has chosen organizational speed over structural separation—a choice that may affect how safety concerns are surfaced relative to deployment urgency.
Pick your industry and the AI tools you use, and get news related to your work every day.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.
In Q2, billionaires or their funds sold six 'Magnificent Seven' AI stocks
Nvidia was rising early Friday, even as the wider chip sector kept falling, Barron's reported
ElevenLabs completed a $300 million employee tender at a $22 billion valuation, and co-founder Mati Staniszews…
On October 15, OpenAI announced the Rosalind Academic Plan, giving approved US university labs free access to…
Neostellar Capital principal Evan Schlossman said CoreWeave's $103.7 billion in remaining performance obligati…
Supermicro Computer appointed Mark Long as chief financial officer, bringing an experienced technology finance…
