Asking artificial intelligence tools about medical symptoms does not help people make better health decisions compared to traditional methods like standard internet searches, according to a new study published in Nature Medicine.
The findings come at a time when a growing number of people are turning to AI chatbots for medical guidance, often assuming these tools provide reliable or safer advice. Researchers caution that despite their popularity, there is still limited evidence that AI health chatbots improve decision-making for patients.
How the Study Was Conducted
The study was led by researchers at the University of Oxford Internet Institute, who collaborated with a group of medical doctors to design 10 different medical scenarios. These scenarios ranged from relatively mild issues, such as symptoms of a common cold, to life-threatening emergencies, including brain bleeding caused by haemorrhage.
Participants were asked to assess these scenarios using different sources of information, including AI-powered health chatbots and conventional online search methods. Researchers then evaluated how accurately participants identified the seriousness of symptoms and whether they chose appropriate actions, such as self-care, visiting a doctor, or seeking emergency help.
Key Findings
The researchers found that people who relied on AI chatbots were no more likely to make correct or safer health decisions than those using traditional internet searches. In some cases, chatbot users underestimated the severity of serious symptoms or felt overly reassured when urgent medical attention was actually required.
The authors noted that AI tools often provide information that sounds confident and authoritative, which can give users a false sense of certainty, even when the advice is incomplete or lacks critical medical context.
Why This Matters
The study’s authors stressed that the findings are especially important as AI chatbots become more accessible and are increasingly marketed as health-support tools. Without proper safeguards, there is a risk that users may delay seeking professional medical care or misinterpret symptoms based on chatbot responses.
Researchers emphasized that AI tools should not replace trained medical professionals and that stronger evidence, regulation, and transparency are needed before such systems can be safely relied upon for health decision-making.
The Bottom Line
While AI health chatbots may offer convenience and quick information, this study suggests they do not improve patient decision-making compared to traditional online searches. Experts continue to advise that individuals consult qualified healthcare professionals when dealing with medical concerns, particularly in cases involving serious or worsening symptoms.
