Everyone uses chatbots to a certain extent now. Some even say the generative AI chatbots have replaced Google search. But when it comes to serious issues such as health information, AI may lead to bigger problems.
A 2025 study in Germany and the UK found that generative AI tools can amplify confirmation bias during health information searches. Researchers used a conceptual and analytical approach to examine how user prompts, AI’s personalized responses, and people’s tendency to favor belief-consistent information interact to reinforce existing views.
The results also indicate that generative AI tools can subtly reinforce confirmation bias by tailoring responses to align with the user’s existing beliefs. This reflective personalization may make belief-consistent answers feel validated, while contradictory information becomes easier to ignore.
Compared with traditional search engines, the study found that conversational AI creates a smoother, more persuasive feedback loop, making it more likely for users to stay within their own viewpoint without realizing it. That said, when you have an assumption about your conditions, or you have already believed it, AI may just reflect what you put in back to you without providing neutral, useful information that you need to know.
Most AI services already state that AI can make mistakes. When you need support in fields such as healthcare, finance, law, or science, double-check information and verify by looking for a reliable source via search engines or licensed professionals.
*What is the Daily Insight? An ongoing series of quick, bite-sized brain snacks. Every week, there are three research-based facts and three research-informed reflective notes.
