Understanding and Designing AI Systems that Reshape Social Dynamics in Group Decision-Making
Abstract
As AI integrates into teams and organizations, it can influence how people communicate, collaborate, and make decisions. I study AI not only as a tool, but as a social actor that can reconfigure group dynamics, power relations, and collective reasoning. My dissertation asks how AI changes group thinking, decision-making, and value exploration, and how we can design it to support more inclusive outcomes. So far, I have completed two studies. First, I conducted a controlled experiment with three LLM agents to investigate the influence patterns of majority, minority, and diffusion over time, and I observed distinct trajectories of opinion change and confidence across tasks. Second, I developed an LLM-powered system that supports minority voices in power-imbalanced group discussions by providing counterarguments or AI-mediated messages. The results show a clear trade-off between participation and psychological safety. Next, I will develop a Value Constellation dashboard that visualizes hidden value conflicts to prompt group reflection and reduce groupthink.
Fields of Study & Subjects
Team Dynamics and Performance
Innovation
Sustainability
Human-Machine Systems
Embodied and Extended Cognition
Publication details
- Journal / source
- Companion Proceedings of the 31st International Conference on Intelligent User Interfaces
- Publisher
- ACM
- Pages
- 234–236
- Publication date
- 9 March 2026