Recent studies from Cornell University and MIT have revealed that 20% reduction in conspiracy theory belief can be achieved by using AI chatbots.

The proliferation of conspiracies on the web has given rise to serious concerns, some of which have caused substantial harm and misunderstanding. According to the latest research from Cornell University and the MIT Sloan School of Management, AI chatbots may be an effective weapon in the fight against these myths. According to a study that was published in Science, conversing with large language models (LLMs) such as GPT-4 Turbo can cut the prevalence of conspiracy theories by about 20%.

In order to evaluate the efficacy of AI chatbots, researchers, including Thomas Costello of MIT Sloan and Dr. Yunhao Zhang from the Psychology of Technology Institute, texted 2,190 participants about their favorite conspiracy theories. The AI was designed to offer solid, evidence-based refutations that were specific to each idea. The study found that participants who engaged with the chatbots significantly reduced their faith in these hypotheses.

A qualified fact-checker examined the assertions made by the chatbot as part of the study to guarantee that it was responding accurately. The accuracy of almost all (99.2%) of these assertions demonstrated the dependability of the data supplied by the AI. According to the research, AI chatbots might be used on several platforms to refute false information and promote users' critical thinking.

Even while the findings seem encouraging, more investigation is required to determine whether chatbots can effectively alter views over time and deal with various kinds of false information. The possibility for incorporating artificial intelligence (AI) into social networking sites and other platforms to improve public education and refute bogus conspiracy theories is highlighted by researchers such as Drs. David G. Rand and Gordon Pennycook.