Here’s what to know before you upload your medical data to an AI program

Key Points:

• A recent KFF poll found that one in three American adults have consulted AI for health advice, with OpenAI’s ChatGPT being widely used by 44% of US adults to answer health questions and interpret personal medical data.
• While AI can analyze uploaded medical records and wearable data to provide personalized health insights, significant privacy concerns remain, as AI companies are not bound by the same strict federal privacy laws that govern hospitals and doctors.
• Experts caution that although AI models like ChatGPT have passed medical licensing exams and can reason at a high level, they still have biases, can provide incorrect or dangerously misleading advice, and may fail to recommend urgent care in emergencies.
• Privacy advocates warn of the risks of sharing sensitive health data with private AI companies without stronger legal protections, citing past data breaches such as the 23andMe incident, and call for improved regulatory frameworks.
• Health professionals recommend using AI tools responsibly by verifying information, asking clear factual questions, starting fresh conversations, and trusting one’s own judgment, especially in urgent medical situations where immediate professional care is necessary.