Evaluating generative AI’s potential to dispel misinformation about wind farms

Abstract Public misinformation about wind farms threatens the global transition to net-zero and a more environmentally sustainable future. This study examines whether conversations with Generative AI (GenAI) can effectively address misinformation and improve attitudes and beliefs about wind farms. In two pre-registered experiments (collective N = 2405), participants with anti-wind farm beliefs engaged in three-round dialogues with ChatGPT, a widely used GenAI tool. Fact-checking showed no clear cases of the GenAI introducing misinformation. Furthermore, following GenAI conversations, participants displayed reduced agreement with misinformation about wind farms, increased policy support, and reduced confidence in their anti-wind farm views. However, some of these effects decayed over time and were not always more effective than static informational resources. These findings highlight both the potential and limitations of GenAI in combating sustainability misinformation, offering insights for leveraging AI in public communication strategies.