Dr Chatbot Will See You Now: The Dangerous Certainty of AI Health Advice

Need a quick diagnosis for a stubborn symptom or a sudden rash? For millions of us, the first instinct is no longer calling the GP—it’s asking an AI chatbot. But a major new study has exposed a terrifying reality behind that instant convenience: up to half of the medical responses generated by leading AI systems contain significant inaccuracies, misleading claims, or flat-out fabrications. Even worse, these bots are designed to sound completely confident, authoritative, and well-referenced, even when their medical citations are entirely made up. Brent is joined by Professor Enrico Coiera, Director of the Centre for Health Informatics at Macquarie University. With more than three decades of research at the intersection of healthcare and information technology, Professor Coiera unpacks why AI algorithms don't actually "understand" medicine, why open-ended symptom searches produce the most dangerous results, and how the absence of clinical context puts everyday users at serious risk. Hit play to find out why AI might be a useful starting point for research, but why relying on it as a source of truth could permanently damage your health.See omnystudio.com/listener for privacy information.