AI chatbots risk steering patients toward unproven cancer therapies, study finds
TL;DR
A study evaluating multiple AI chatbots (Gemini, DeepSeek, Meta AI, ChatGPT, Grok) found roughly half produced problematic medical responses when asked about cancer and related health topics, with some suggesting alternative therapies or clinics. Experts warn these flawed replies can mislead patients and stress the need for safer, more reliable AI medical guidance.
Topics:businesshealth#ai-chatbots#alternative-therapies#bmj-open#cancer-treatment#health#health-misinformation
- AI chatbots are giving people alternatives to chemotherapy, study finds NBC News
- Should you really trust health advice from an AI chatbot? BBC
- Chatbots giving dangerous and inaccurate medical advice study finds The Independent
- Public use of a generalist LLM chatbot for health queries Nature
- Americans Ditch Dr. Google For AI Advice Let's Data Science
Want the full story? Read the original reporting
Read on NBC News