A new Anti-Defamation League (ADL)-supported study indicated that interactions with an AI chatbot decreased belief in antisemitic conspiracy theories, according to findings released on Thursday.
The peer-reviewed study, titled "Short Dialogues with AI Reduce Belief in Antisemitic Conspiracy Theories," was conducted by researchers at the ADL’s Center for Antisemitism Research. The research involved participants who expressed belief in specific antisemitic conspiracy theories, such as Holocaust denial and claims about Jewish control of media and government, discussing the theories with an AI chatbot programmed to debunk the ideas.
Researchers found that the interaction reduced their belief in the theories by 16%. The study also found that in participants who held an unfavorable view of the Jewish people, the chatbot intervention led to a 25% increase in favorability towards Jews.
Factual debunking can reduce belief in antisemitic conspiracies
The decreased belief in conspiracy and increased favorability persisted in the month following the initial research period, with researchers reporting that 50% of the results were still evident in follow-up surveys with participants.
According to the Center for Antisemitism Research, the results suggest that “factual debunking can reduce belief in antisemitic conspiracies,” and that the theories’ “deep historical roots and ties to ethnic and religious prejudice do not protect against the power of facts and evidence.”
Jonathan A. Greenblatt, the CEO and National Director of the ADL, expressed that his organization is “constantly innovating to find new ways to counter this age-old hatred with modern technology,” continuing that it is “very encouraging to see that even deeply entrenched prejudice can be effectively challenged by promising new tools that could reach millions at scale.”
David G. Rand, Professor of Information Science, Marketing, and Psychology at Cornell University, and the study's senior author, explained how the chatbot was trained. “Our artificial intelligence debunker bot typically doesn't rely on emotional appeals, empathy-building exercises, or anti-bias tactics to correct false beliefs,” Rand said, “It mostly provides accurate information and evidence-based counterarguments, demonstrating that facts still matter in changing minds.”
While the results appeared promising, researchers emphasized that a participant’s willingness to engage with the AI is the main challenge to achieving broader reductions in conspiratorial beliefs and the correction of misconceptions.