A patient in her 50s sought medical attention after experiencing flashes of light in one eye. The patient presented to an eye clinic, stating she had utilized ChatGPT to determine her condition was migraine with visual aura.

An examination performed at the clinic revealed the patient had a retinal tear. The patient received laser treatment for the retinal tear on the same day as the examination.

The patient's initial symptoms included brief, recurrent flashes of light in her peripheral vision. These flashes were particularly noticeable in dark environments.

Why It Matters

This case occurred as studies examine the use of artificial intelligence in healthcare. A 2026 KFF poll indicated that 32% of U.S. adults reported using AI chatbots for health-related information within the past year. The same poll noted that many of these respondents did not subsequently consult a clinician after seeking health information from AI chatbots.

Research has explored the impact of AI on information processing. A 2024 study conducted by MIT and the University of California-Irvine found that individuals who discussed an event with a generative chatbot developed over three times more false memories compared to a control group. This study also observed that participants maintained elevated confidence in these false memories a week later. Furthermore, a randomized clinical vignette study published in JAMA indicated that clinicians' accuracy decreased when they were presented with biased AI diagnostic predictions, even when explanations from the AI model were provided. A 2026 Nature Medicine study found that while large language models performed well when tested independently, members of the public using these models were no more effective than control groups at identifying relevant medical conditions or selecting appropriate courses of action.