Leveraging AI for democratic discourse: Chat interventions can improve online political conversations at scale

Political discourse is the soul of democracy, but misunderstanding and conflict can fester in divisive conversations. The widespread shift to online discourse exacerbates many of these problems and corrodes the capacity of diverse societies to cooperate in solving social problems. Scholars and civil society groups promote interventions that make conversations less divisive or more productive, but scaling these efforts to online discourse is challenging. We conduct a large-scale experiment that demonstrates how online conversations about divisive topics can be improved with AI tools. Specifically, we employ a large language model to make real-time, evidence-based recommendations intended to improve participants' perception of feeling understood. These interventions improve reported conversation quality, promote democratic reciprocity, and improve the tone, without systematically changing the content of the conversation or moving people's policy attitudes.

Automated summary

Political discourse is crucial for democracy, but misunderstandings and conflicts can arise. The shift to online discourse has worsened these issues, undermining the ability of diverse societies to collaborate in solving social problems. Scholars and civil society groups have proposed interventions to reduce divisiveness and increase productivity, but this poses challenges for online discourse. We conducted a large-scale experiment that demonstrates how AI tools can improve online conversations about divisive topics. By using a large language model, we provide evidence-based recommendations in real-time to enhance participants' perception of being understood. These interventions improve conversation quality, promote democratic reciprocity, and improve the tone, while not altering the conversation's content or people's policy attitudes.