NPR and NewsGuard find chatbots outperform search engines on misinformation
A new study from NPR and NewsGuard shows that AI chatbots, like ChatGPT and Google's Gemini, are much better at catching misinformation than regular search engines.
Researchers tested them with tricky questions based on false claims spread by state actors like China, Russia, and Iran between December 2025 and July 2026.
The results came out today.
Chatbots debunk about 75% false claims
Chatbots managed to debunk about 75% of the false information, while search engines often missed the mark or showed misleading answers.
Google's AI Overview was the most reliable among search engine summaries; Bing struggled the most, with DuckDuckGo somewhere in between.
Experts say it's smart to double-check chatbot answers with primary sources, especially when dealing with news influenced by state-controlled media.