A new study has found that turning to artificial intelligence (AI) tools for medical advice may not give patients better results than using standard health information sources such as internet searches or official websites. Researchers in the UK tested how people used AI chatbots and found that, in real‑world scenarios, these tools did not improve decision‑making for common medical symptoms.
The study, published in Nature Medicine and led by the University of Oxford’s Internet Institute, involved nearly 1,300 participants from Britain who were presented with various medical situations, ranging from minor ailments to more severe health concerns. While advanced AI models like ChatGPT‑4o, Llama 3, and Command R+ correctly identified medical conditions in controlled experiments, those same models did not help people choose the right actions in practical use better than traditional resources.
According to the researchers, the gap between AI’s technical potential and actual effectiveness may be due to how people communicate their symptoms and how the systems interpret them. As AI becomes more widely accessible, the findings highlight that users should be cautious about relying on chatbot responses for health decisions and continue to seek verified medical information and professional advice.
Source: Reuters – AI no better than other methods for patients seeking medical advice, study shows (https://www.reuters.com/business/healthcare-pharmaceuticals/ai-no-better-than-other-methods-patients-seeking-medical-advice-study-shows-2026-02-09/)