Mitigating Conflicts in Communication Between Opposing Positions Through the Intervention of Generative AI

Abstract

The objective of this study is to examine the impact of a dialogue intervention system using generative AI on the communication between individuals with opposing views. Specifically, in debates conducted via online chat tools, participants' messages were filtered and modified by generative AI to eliminate aggression and rewrite them with empathetic content. This system aims to reduce the psychological burden in the exchange of opinions between hostile parties by adding empathetic expressions. In the experiment, a debate was conducted between two individuals with conflicting opinions, with the exchange of messages facilitated by generative AI. The survey results obtained after the experiment indicated a tendency for reduced psychological burden on the participants due to the intervention of the system. This suggests the potential of improving interpersonal relations in communications between two opposing parties through the intervention of generative AI.