This article summarizes a study examining how people assess medical advice generated by large language models compared to that from medical doctors. The researchers found that 300 participants often could not tell the difference between AI-generated and doctor-written responses, displaying a preference for AI advice—rating it as more trustworthy, valid, and complete even when accuracy was low. Both experts and nonexperts showed bias toward AI-generated content and were more likely to act on inaccurate AI recommendations, potentially leading to harmful outcomes. The findings highlight the risk of overtrust in AI for healthcare, stressing the need for careful integration of AI with human oversight to prevent the spread of medical misinformation.
Related articles:
How Artificial Intelligence Is Changing Medical Practice
The Need for Human Oversight in AI-Driven Medical Decisions
Medical Errors Associated With New Technology
AI in Health Care: Balancing Innovation and Patient Safety
How to Make Medical AI Trustworthy
