OpenAI is reorganising the research team responsible for shaping the behaviour of its AI models, including ChatGPT. This adjustment comes as the company addresses user feedback on model interactions, particularly following the introduction of GPT-5 and its new personality options. The team's leader is transitioning to a different project within OpenAI.
The reorganisation aims to streamline efforts in refining AI personality, reducing sycophancy, and enhancing features like emotional detection in chats. OpenAI is also focused on integrating behavioural research with emerging technologies like AI agents. This shift follows adjustments such as the restoration of GPT-4o after user dissatisfaction with GPT-5's initial personality tweaks.
This move positions OpenAI to refine AI personality and address ethical concerns. The company is integrating GitHub connectors for code analysis in ChatGPT's deep research tools, betting on a more integrated approach where behavioural traits support complex tasks.
Related Articles
GPT-5: Proceed with Caution
Read more about GPT-5: Proceed with Caution →
GPT-5 Gets Friendlier Update
Read more about GPT-5 Gets Friendlier Update →
ChatGPT Model Selection Returns
Read more about ChatGPT Model Selection Returns →
ChatGPT Adds Third-Party Connectors
Read more about ChatGPT Adds Third-Party Connectors →