Trust and Value in Negotiation: Comparing AI and Human Counterparts

Abstract: Advances in artificial intelligence (AI), particularly in large language models, are transforming negotiation processes. This study investigates how human versus AI counterparts influence economic and relational negotiation outcomes. In an experimental study with 205 participants negotiating apartment rent, participants in the Person condition paid significantly higher rent but reported greater relational satisfaction, including higher trust, warmth, and perceived competence, compared to the AI condition. Exploratory factor analysis revealed distinct evaluation structures: human counterparts prompted nuanced perceptions across relational, procedural, and competence dimensions, while AI counterparts elicited unified evaluations. These findings highlight a trade-off between relational engagement and economic efficiency in negotiations. While AI enables more favorable economic outcomes, human counterparts foster trust and relational satisfaction. This work provides practical implications for designing AI systems that balance efficiency with relational engagement, and it encourages further research on human-AI dynamics in negotiations.

Keywords: Negotiation, Artificial Intelligence, Relational Dynamics, experiment