Medical experts caution against taking health advice from AI chatbots, following a study that revealed these systems, including ChatGPT, suggest harmful practices like inserting garlic into the rectum for immune support. Researchers found that when misinformation was presented in clinical language, AI models were more likely to endorse false medical claims, including dangerous recommendations. The findings highlight the unreliability of AI for health decisions, stressing the need for expert oversight in AI applications in healthcare.
Want More Context? 🔎