ChatGPT Health Launches to Help Decode Your Health Data

OpenAI has officially launched Health in ChatGPT for users in the United States. This feature allows users to connect Apple Health and supported medical records so ChatGPT can help make sense of health information with context, track changes over time, and deliver more personalized conversations. The initiative responds to the needs of over 300 million weekly users who often ask health-related questions, yet the context behind those questions is scattered across different apps and patient portals.

How Health in ChatGPT Untangles Scattered Health Data

With user permission, ChatGPT can now leverage connected Health information to compare new test results with prior records, summarize changes since the last visit, or explore relationships between sleep, activity, and exercise routines. This reduces the need to repeatedly upload or explain the same details, while helping users feel more informed and proactive in managing their own health, without replacing the role of medical professionals.

Notably, health conversations are no longer confined to a dedicated space. Early testing showed over 70% of health-related chats occurred outside the separate Health experience. Based on that feedback, users can now let ChatGPT consider connected health data across all conversations. For example, when picking a restaurant, ChatGPT can keep dietary restrictions in mind, or when planning weekend activities, it can adjust for a recent injury. Meanwhile, the Health sidebar remains the hub for managing information, viewing trends, and browsing health chat history.

"Thanks to Health, my scattered medical history now makes sense. ChatGPT connects various diagnoses, imaging results, and surgery notes into a clear big picture, and explains medical terms in simple language. I feel more confident during doctor or therapist visits."
- Blake, technical program manager

AI Models Rigorously Tested with Hundreds of Doctors

OpenAI continues to improve the models behind health conversations. GPT-5.5 Instant, available to free-tier users, is now more reliable at recognizing emergency situations, requesting missing context, and explaining uncertainty. On the toughest health evaluations, its performance matches that of frontier reasoning models at the time. Meanwhile, GPT-5.6 Sol, available to paid users, offers stronger abilities in reasoning over complex details, explaining doctor visit notes, and preparing questions for follow-up appointments.

To measure progress, OpenAI collaborated with hundreds of doctors worldwide to design realistic health scenarios and rigorous evaluation rubrics. These evaluations assess accuracy, safety, communication, context awareness, completeness, and appropriate escalation. Every GPT-5.6 model outperformed GPT-5.5 on HealthBench Professional, a benchmark for AI performance on complex health tasks. Before release, doctors also tested Health in ChatGPT hands-on to ensure real-world safety and performance. Still, OpenAI reminds users that ChatGPT may make mistakes and does not replace the judgment and care of medical professionals.

Privacy and Control: Users Hold the Reins

Health is built with layered privacy protections. All ChatGPT conversations are encrypted in storage and transit, and Health-connected information receives additional encryption safeguards. Importantly, linked medical record and Apple Health data is not used to train foundation models or target ads, regardless of the user's model training settings. Conversations that do not draw on this data follow the user's standard settings.

Users have full control: by default, ChatGPT will ask for permission before using health data to personalize responses. This permission can be granted one time or set to always allow. Users can also disconnect their accounts at any time, and synced data will be deleted from OpenAI's systems within 30 days. Memories formed from health conversations are fully manageable, and chats can take place in Temporary Chat mode to leave no trace.

What This Means for Users and Developers in Indonesia

Though currently available only in the U.S., the arrival of Health in ChatGPT signals a direction for integrating personal health data with increasingly smart and safe AI assistants. For developers in Indonesia, the strict privacy approach, transparent user controls, and collaboration with medical professionals can serve as a blueprint for building similar solutions in the future. For users, it paints a picture of how AI might one day help organize complex health information, while keeping doctors as the primary partners in medical decision-making.