What happens when AI agents move beyond assisting individuals and begin helping groups negotiate, resolve conflict, and reach agreement? In this lightning talk, I introduced my research interest in multi-agent negotiation, exploring how LLM-based agents could identify common ground, reduce mistrust and miscommunication, and support more constructive collective decision-making. Using Concordia as a social simulation framework, the work investigates how different agent strategies may improve cooperation across complex negotiation scenarios.
Links:
- The workshop: Citizen-Centric Artificial Intelligence Systems Workshop 2025
- My lightning talk slide: Multi-Party Negotiation Lightning Talk Slides