Read, Watch or Listen

OpenAI is restructuring its Model Behavior team, merging it with the Post Training team to integrate AI personality development more closely with model creation. The team, known for mitigating sycophancy and bias in models like GPT-4 and GPT-5, will now report to Max Schwarzer. Team founder Joanne Jang is launching OAI Labs, focusing on new AI interaction interfaces beyond the current chat paradigm. This reorganization follows recent criticism and a lawsuit related to ChatGPT's responses. Jang's new venture aims to explore AI's role as an instrument for various creative and collaborative tasks.

Prepared by Jonathan Pierce and reviewed by editorial team.