OpenAI is reshaping one of its most influential research units as part of a wider effort to refine how its artificial intelligence systems interact with users, according to a TechCrunch report.
The publication highlighted that in an internal memo circulated in August, chief research officer Mark Chen told staff that the company’s Model Behaviour team would be integrated into the larger Post Training division.
This move will reportedly see the 14-member group, which has played a pivotal role in shaping how OpenAI’s models communicate with people, report directly to Max Schwarzer, who leads the Post Training team.
Notably, the Model Behaviour unit has been responsible for defining the “personality” of OpenAI’s AI systems. Its projects have included reducing sycophancy, where models agree uncritically with users, and managing political bias in responses.
Reportedly, the group also contributed to OpenAI’s internal discussions about whether AI systems can or should be considered conscious. Chen said the decision to restructure would ensure this work is embedded at the centre of model development, reflecting the company’s view that personality is now a critical factor in how AI evolves, adds the report.
The founding head of the team, Joanne Jang, will not be part of the reorganisation. Instead, she is establishing a new initiative within the company called OAI Labs. This new unit will be responsible for exploring and prototyping novel ways for people to collaborate with AI, with the aim of making interactions more intuitive and productive, noted the report.
At the same time, OpenAI has confirmed plans to launch a new jobs platform in 2026. The service will connect employers with candidates who possess skills in artificial intelligence, amid growing demand for expertise in areas such as machine learning, data analysis and automation. The move also comes against a backdrop of concern over how AI could reshape employment in sectors already facing disruption.
