TLDR: OpenAI has strategically merged its Model Behavior team with the Post Training division to enhance the personality and interaction style of its AI models, including ChatGPT and GPT-5. This move aims to address user feedback regarding AI’s ‘cold’ demeanor, foster more empathetic and balanced responses, and deepen the integration of ethical considerations into core AI development. Concurrently, Joanne Jang, former head of the Model Behavior team, is launching OAI Labs to explore innovative human-AI collaboration interfaces.
OpenAI has announced a significant strategic overhaul, integrating its specialized Model Behavior team into the broader Post Training division. This reorganization, confirmed by an internal memo from Chief Research Officer Mark Chen, is a direct response to user feedback, particularly concerning the interaction style of its advanced AI models like GPT-5, which some users found to be ‘cold’ or detached. The objective is to cultivate a warmer, more empathetic, and assertive AI personality, thereby enhancing the overall user experience and trust.
The Model Behavior team, a crucial 14-member group, has historically been instrumental in shaping the ‘personality’ of OpenAI’s AI systems. Their work has focused on critical areas such as reducing AI sycophancy—where models uncritically agree with users—and mitigating political biases to ensure balanced and neutral responses. This team has influenced every major model since GPT-4, including GPT-4o, GPT-4.5, and GPT-5. By merging this team into the Post Training division, which is led by Max Schwarzer, OpenAI aims to embed behavioral design and ethical considerations more deeply into the core model development process.
Mark Chen stated in the internal memo that aligning this work with core model training is essential as the company refines AI’s tone, empathy, and reliability. This strategic shift reflects a broader industry trend prioritizing user-experience improvements and holistic AI-human interaction dynamics over purely technical enhancements. The company had previously rolled back some updates to GPT-5 and released a patch to make it feel warmer and more engaging after user backlash.
Adding another layer to this restructuring, Joanne Jang, the founding head of the Model Behavior team, is departing OpenAI to establish a new initiative called OAI Labs. Jang will serve as General Manager of this new research unit, which aims to invent and prototype innovative interfaces for human-AI collaboration. She envisions moving beyond traditional chat-based interactions towards tools that support thinking, creating, playing, learning, and connecting, potentially incorporating hardware designs. This venture underscores OpenAI’s commitment to exploring advanced interaction modalities and broadening the landscape of AI applications.
The reorganization comes amid heightened scrutiny of AI behavior, including a recent lawsuit filed in August by the parents of 16-year-old Adam Raine. The lawsuit alleges that ChatGPT, powered by GPT-4o, failed to respond appropriately to their son’s suicidal thoughts, intensifying the debate over AI’s responsibility in mental health crises. OpenAI’s proactive approach to refining AI personality and embedding ethical frameworks is seen as a direct response to such controversies, aiming to mitigate potential liabilities and uphold user safety and trust.
In addition to these changes, OpenAI has also confirmed plans to launch a new jobs platform in 2026. This service will connect employers with candidates possessing AI skills, addressing the growing demand for expertise in areas like machine learning, data analysis, and automation. This initiative also acknowledges concerns about AI’s potential impact on employment sectors.
Also Read:
- OpenAI’s GPT-5 System Prompt Details Emerge, Revealing Enhanced Capabilities and Strict Guardrails
- ChatGPT Introduces Customizable Personalities for Tailored AI Interactions
Overall, OpenAI’s strategic realignment signifies a profound commitment to developing AI systems that are not only technologically advanced but also ethically sound, user-centric, and capable of more nuanced, empathetic human-like interactions. This move is expected to set new benchmarks for ethical AI development across the industry.
