AI Chatbots Telling Cancer Patients to Try Useless Woo-Woo Treatments Instead of Chemotherapy
A new study found that leading AI chatbots often provide unscientific and potentially harmful medical advice, including recommending unproven cancer treatments. The chatbots' responses were deemed 'problematic' half of the time, with some suggesting alternative therapies over chemotherapy.
Researchers tested the accuracy of free AI chatbots, including OpenAI's ChatGPT and Google's Gemini, on health topics prone to misinformation. The chatbots provided 'problematic' responses half of the time, with 20% being 'highly problematic'. When asked about cancer treatments, the chatbots sometimes suggested alternative therapies like acupuncture and herbal medicine alongside chemotherapy, presenting unscientific claims on equal footing with scientific ones. This 'false balance' could lead cancer patients to forgo necessary medical help. The study's findings are concerning, given that tens of millions of Americans use AI chatbots for health advice. The chatbots' inability to provide clear, science-based answers is a fundamental flaw that needs to be addressed.
This content was automatically generated and/or translated by AI. It may contain inaccuracies. Please refer to the original sources for verification.