In recent years, there has been a significant surge in the use of artificial intelligence for information gathering, particularly with large language models (LLMs) like ChatGPT, Claude, and Gemini being increasingly utilized for news verification and consumption. According to reports from the Pew Research Center, one in five U.S. teens regularly uses LLMs to get news, while one in four young adults has used them at least once. A new open-access study from the MIT Media Lab reveals potential dangers of this trend: Researchers found that participants who relied on AI systems to verify facts actually became worse at detecting misinformation on their own once their chatbots were removed, a phenomenon termed the “AI dependency paradox.”
The study tracked 67 participants over four weeks as they evaluated news headline-image pairs. Participants were 21% more accurate in detecting fake news when assisted by an AI chatbot during a session, confirming previous research from MIT Sloan School of Management that AI can effectively reduce people's belief in false information. However, the study showed that a new wrinkle emerged when the AI was no longer present: by week four, participants’ unassisted performance on new news items declined by 15 percentage points compared to before the study started. Roughly a quarter of all participants reported feeling that they were getting better at detection, even as their performance declined.
The qualitative analysis identified distinct behavioral patterns, with one-fifth of participants labeled as