Recent research suggests that AI, like the GPT series, can effectively reduce harmful conspiracy beliefs by engaging believers in personalized dialogues. This approach challenges their views with fact-based counterarguments, leading to a significant reduction in conspiracy belief that persists long-term.