OpenAI is restructuring its Model Behavior team, a group of researchers who influence how the company’s AI models interact with users, according to a memo obtained by DailyTech.
The Model Behavior team, consisting of around 14 researchers, will now be part of the larger Post Training team, which focuses on enhancing AI models after their initial training phase.
Under the new structure, the Model Behavior team will report to Max Schwarzer, the lead of OpenAI’s Post Training team. A spokesperson for OpenAI confirmed these changes.
Joanne Jang, the leader of the Model Behavior team, is moving on to launch OAI Labs, a research team dedicated to developing new interfaces for human-AI collaboration.
The Model Behavior team plays a crucial role in shaping the personality of OpenAI’s AI models and ensuring they provide balanced responses without reinforcing negative beliefs. They have also addressed issues like political bias and AI consciousness.
OpenAI is emphasizing the importance of the AI’s “personality” by integrating the work of the Model Behavior team more closely with core model development.
In response to user feedback, OpenAI made adjustments to its GPT-5 model to strike a balance between friendliness and avoiding sycophancy. The company faced criticism over the behavior of its AI models, particularly in cases like the ChatGPT controversy.
Techcrunch event
San Francisco
|
October 27-29, 2025
The Model Behavior team has been involved in the development of every OpenAI model since GPT-4, including versions like GPT-4.5 and GPT-5. Jang, the former team lead, has also worked on projects such as Dall-E 2.
Jang will now head OAI Labs, a new initiative within OpenAI focused on exploring innovative interfaces for human-AI interaction. She envisions moving beyond traditional chat-based interactions to create new tools for thinking, creating, and learning.
When asked about potential collaborations with Jony Ive on AI hardware devices, Jang expressed openness to various ideas but plans to focus on familiar research areas initially.
Update 4:35pm PT: This story was updated to include a link to Jang’s post announcing her new position, which was released after this story published. We also clarify the models that OpenAI’s Model Behavior team worked on.
