OpenAI has announced the launch of a new feature designed specifically for the health and wellness industry - ChatGPT Health. This is the first independent interaction space built by ChatGPT for a specific privacy-sensitive field. The purpose of this feature is to solve the pain point of scattered and difficult-to-reuse user health data, helping people manage their health indicators more efficiently.
Currently, over 230 million users worldwide try to seek health advice through ChatGPT each week. To provide more accurate and personalized feedback, ChatGPT Health allows users to securely connect their electronic health records (EHR) and mainstream health apps, including Apple Health, MyFitnessPal, and Function, among others. This means that ChatGPT can help you interpret test reports based on your real lab results, sleep data, or diet habits, and prepare targeted questions for your upcoming doctor's visit.
Privacy protection is the core of this upgrade. OpenAI has established a completely isolated "health" space, where the conversation data within this space is not only encrypted with dedicated technology but also clearly committed to not being used for training its base model. In addition, this space has an independent memory system, ensuring that health information does not leak into the user's regular chat conversations.
To ensure the professionalism of the information, OpenAI has collaborated with over 260 practicing doctors around the world for the past two years, conducting more than 600,000 feedback trainings on the model. The model's responses follow the HealthBench evaluation framework written by doctors, prioritizing safety and clinical standards. It should be noted that this feature is positioned as a medical assistant tool rather than a diagnostic tool, with the core goal of complementing but not replacing professional clinical care. Currently, this feature is being gradually rolled out to some users, and users can join the waiting list via the web or iOS app.
Key Points:
🔐 Data Isolation and Enhanced Privacy: A separate "health" conversation space is set up, using dedicated encryption technology, and related conversation data will never be used for AI model training.
🏥 Deep Integration with the Health Ecosystem: Supports connecting to electronic health records (EHR) and third-party applications such as Apple Health, allowing AI to provide advice based on real lab results and fitness data.
🩺 Deep Optimization by Doctors: Collaborated with over 260 doctors worldwide for clinical training, ensuring that AI responses are logically and safely aligned with medical standards in terms of professional explanations and medical advice.
