Overview
- OpenAI added GPT Health inside ChatGPT to let users link health and wellness apps and upload medical records so the chatbot can use that data to give personalized answers.
- OpenAI says uploaded health information will not be used to train models or to target ads, a privacy assurance meant to ease user concerns about data use.
- Doctors and projects such as YouEatSafely describe concrete, nonclinical uses for ChatGPT including reading food labels, analyzing meals, improving health literacy, and translating guidance into other languages.
- A published study led by Dr. Nick Tiller found chatbots gave problematic or highly problematic medical responses a third to half the time and in many tested emergencies advised delaying care for 24 to 48 hours, and he notes those tests used older models and that current models lack public, independent audits.
- The rollout increases both benefit and risk by making personalized AI health advice easier to access, which could deepen public reliance but also create safety hazards without clinical validation, transparent audits, or clearer regulatory guidance.