When the AI model was instructed to persuade users without using facts or evidence, its debunking effect on conspiracy beliefs was completely eliminated.
Sources
- Chatbots are surprisingly effective at debunking conspiracy theories (technologyreview.com)
- https://www.science.org/content/article/ai-chatbot-shows-promise-talking-people-out-conspiracy-theories (science.org)
- https://mitsloan.mit.edu/ideas-made-to-matter/mit-study-ai-chatbot-can-reduce-belief-conspiracy-theories (mitsloan.mit.edu)
- https://arxiv.org/pdf/2510.01537 (arxiv.org)
- https://arxiv.org/pdf/2506.09292 (arxiv.org)
More in Psychology
- In a study published in Science, more than 2,000 conspiracy believers had an approximately 8.4-minute, three-round conversation with an AI chatbot called DebunkBot, which reduced their confidence in the conspiracy by 20%.
- Researchers found that human imagination's capacity for tracking disappeared moving objects was limited to one object, contrary to their initial expectation that it could track approximately three or four.
- A 1997 study by E.N. Aron and A. Aron established a link between sensory-processing sensitivity and traits of introversion and emotionality, published in the Journal of Personality and Social Psychology.
- In 2012, Max Bazerman and four coauthors published a paper claiming that signing a promise to tell the truth before filling out a form produced greater honesty than signing it afterward.
- In the Open Science Collaboration's 2015 replication project, only about 36% of 100 replicated psychology studies produced statistically significant results similar to the originals.
- In Greene and Paxton's study, participants who behaved mostly honestly showed the same reaction times whether they had an opportunity to cheat or not, and whether they won or lost money.