New research from Cornell, MIT, and American University has demonstrated that AI-powered dialogues can effectively reduce belief in conspiracy theories. The study, published Sept. 13 in Science, involved over 2,000 participants engaging with OpenAI’s GPT-4 Turbo, which tailored personalized, evidence-based responses to debunk various conspiracies.
The study found that, on average, belief in conspiracy theories dropped by 20% after these AI-driven conversations. Notably, one in four participants disavowed their conspiracy beliefs entirely. The research highlights AI’s potential to combat misinformation and reshape public discourse on contentious issues like COVID-19 and election fraud.
Researchers emphasized the importance of responsible AI use, as the technology could influence beliefs both positively and negatively. However, the study marks a promising step toward using AI to counter the spread of conspiracy theories at scale.
FingerLakes1.com is the region’s leading all-digital news publication. The company was founded in 1998 and has been keeping residents informed for more than two decades. Have a lead? Send it to news@fingerlakes1.com.