3 strategies for responding to patients who use AI chatbots for medical advice

Patients increasingly arrive at appointments having already consulted an AI chatbot about their symptoms, medications or diagnosis. The trend has grown quickly. A recent KFF poll found that about one in three adults used AI chatbots for health information over the past year, which puts AI on par with social media as a source for medical questions, as Medical Economics recently reported. Use skews toward younger adults, and among those who used AI for physical health advice, 42% never followed up with a clinician.

For physicians, these conversations bring both opportunity and risk. A patient who has researched a condition may be more engaged and better prepared to ask questions. But chatbot answers aren't always reliable. Researchers at Mount Sinai found that chatbots fabricated diseases, lab values and clinical signs in up to 83% of simulated cases when no safeguards were in place, and the models tend to repeat and build on false details users feed them.

Dismissing a patient's AI research can damage trust. Accepting it uncritically can lead to poor decisions. Physicians need a middle path: acknowledge the effort, correct misinformation clearly, and position themselves as the trusted interpreter of what patients find online.

In this episode of Vital Signs, Dr. Sarah Matt describes three strategies to respond to patients who already have an AI-generated answer to their problem.