Abstract This article discusses the projects presented on the “AI in the Field” panel at the March 2025 AI Negotiation Summit. The projects addressed distinct domains of social conflict: family caregiving, humanitarian frontline negotiations, and democratic deliberation. In contrast to current discourse surrounding autonomous “AI agents” intended to replace human roles, the research presented in this panel explicitly focused on human–AI collaboration. Across all three studies, participants engaged productively with AI systems designed to support decision-making and conflict management. The deliberative democracy project showed that AI-supported deliberation could promote opinion change and movement toward common ground on some divisive sociopolitical issues. The frontline negotiators study demonstrated that a structured AI system improved efficiency compared with negotiators self-prompting ChatGPT; its interactive interface enabled users to pose questions, generate multiple options, and assess risks across several dimensions, supporting deeper analysis and counterfactual reasoning. The caregivers study showed evidence of learning, behavioral change, and use in social conflict beyond the training context. Each project also addressed the risk of negative social consequences associated with AI systems by using multidisciplinary teams, grounding interventions in established theory, conducting extensive formative research, and maintaining human oversight of system behavior. Together, the studies provide proof-of-concept evidence that carefully designed AI systems can enhance human capacity to address social conflict while preserving human judgment and minimizing the risks of human dependency and unintended social harm.
Jeanne M. Brett (Thu,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: