A new study has found that artificial intelligence chatbots misdiagnose medical conditions in more than 80% of early cases, raising concerns about the growing use of AI tools in healthcare decision-making.

The research comes as patients increasingly turn to AI-powered chatbots for medical advice and preliminary diagnoses. The high error rate documented in the study suggests significant limitations in current AI technology when applied to complex medical assessments.

The findings have personal resonance for some researchers in the field. Ben Riley, who has written about the risks of AI chatbots, experienced firsthand the potential dangers when his father began trusting artificial intelligence recommendations over advice from his doctor.

The case highlights a broader trend of patients seeking medical guidance from AI systems, often without understanding their limitations. Medical professionals have expressed growing concern about individuals making healthcare decisions based on chatbot recommendations rather than consulting qualified physicians.

The study's results underscore the need for caution when using AI tools for medical purposes, particularly given the potential consequences of misdiagnosis. Healthcare experts emphasize that while AI may serve as a useful supplementary tool, it should not replace professional medical consultation and diagnosis.