OpenAI has rolled out a new US-only ChatGPT feature that analyses users’ medical records and data from apps like MyFitnessPal to deliver more personalised health advice. Privacy campaigners have voiced serious concerns.
Conversations in ChatGPT Health are stored separately and will not be used to train AI models. OpenAI stresses the tool is not for diagnosis or treatment.
Andrew Crawford of the Center for Democracy and Technology called for “airtight” safeguards around health data, noting it is among the most sensitive personal information. He warned that as OpenAI explores advertising, strict separation from other ChatGPT data is essential.
OpenAI says over 230 million people ask health-related questions weekly. The feature includes “enhanced privacy” protections and supports data from Apple Health, Peloton, MyFitnessPal and uploaded records to improve response relevance. It is designed to support—not replace—professional medical care. OpenAI has assured users that all conversations held within ChatGPT Health will be kept separate from other chats and will not be used to train its AI models. The firm has also made clear that the feature is not designed to provide medical diagnosis or treatment.
Max Sinclair of Azoma called the launch a “watershed moment” that could reshape patient care and retail by positioning ChatGPT as a trusted health adviser and influencing consumer purchases. He described it as a potential “game-changer” amid competition from Google’s Gemini.
The feature is initially available to a small group of early users via waitlist only. It is not available in the UK, Switzerland or the European Economic Area due to strict data protection rules.
Author: George Nathan Dulnuan