ChatGPT Health rollout raises concerns among US medical community

OpenAI has begun rolling out ChatGPT Health, a feature that lets U.S. users aged 18 and older connect personal medical records, wearables and wellness apps to receive AI‑generated health information. The service is available on web and iOS and is being introduced across all subscription tiers, though the rollout may take weeks to reach all users.

Healthcare professionals warn that patients might treat the AI’s responses as medical advice, creating privacy, consent and data‑security issues, and potentially altering the patient‑provider relationship. At the same time, recent scientific studies have produced conflicting assessments of AI performance in clinical settings. A Nature Medicine paper suggested that general‑purpose models such as ChatGPT, Claude and Gemini outperformed a specialized model, OpenEvidence, while a later JAMA study found the opposite, rating OpenEvidence higher across accuracy, utility and verifiability. The contradictory findings highlight the need for independent, transparent evaluation of AI tools before they are widely adopted in medical practice.