ChatGPT Health marks OpenAI’s most serious move into healthcare yet, offering a privacy-first, physician-informed AI experience designed to help users understand their health—without replacing clinical care. (Source: Image by RR)

Physician Collaboration Shapes Safety and Response Quality

OpenAI has announced ChatGPT Health, a dedicated experience designed to help people better understand and manage their health by securely combining personal health information with ChatGPT’s intelligence. Health is already one of the most common use cases for ChatGPT, with more than 230 million people globally asking health and wellness questions each week. The new experience, as noted at openai.com, formalizes that behavior into a protected, purpose-built environment focused on clarity, preparedness, and user confidence.

ChatGPT Health addresses a familiar problem: health data scattered across portals, PDFs, wearables, and apps, leaving individuals to piece together meaning on their own. Users can now securely connect medical records and wellness platforms—including Apple Health, MyFitnessPal, Function, and others—so conversations are grounded in personal context. The tool helps users interpret lab results, prepare for doctor visits, understand trends over time, and navigate lifestyle or insurance questions, while explicitly not offering diagnosis or treatment.

Privacy and security are central to the design. Health operates as a separate, compartmentalized space within ChatGPT, with dedicated memories, added encryption, and strict isolation from non-health chats. Health conversations are not used to train OpenAI’s foundation models, and users maintain full control over connected data, app permissions, and stored memories. Purpose-built encryption, optional multi-factor authentication, and partnerships with secure health data providers like b.well reinforce protections for sensitive information.

The experience was developed in close collaboration with physicians worldwide, with over 260 doctors across 60 countries contributing feedback on more than 600,000 model evaluations. This work informed both response style and safety thresholds, guided by HealthBench, OpenAI’s clinician-designed evaluation framework. Access is rolling out gradually via a waitlist, with plans to expand availability on web and iOS. As OpenAI positions ChatGPT Health as a companion—not a replacement—for medical care, the launch signals a broader push toward regulated, trust-centered AI in healthcare.

read more at openai.com