ChatGPT Health rolls out widely to US adults, but Android users face platform gaps

OpenAI has started a wide rollout of ChatGPT Health to all US adults aged 18 and older, moving the feature out of its waitlist phase. The tool lets users connect health data sources such as Apple Health, One Medical, Function Health, and supported US hospital system records to receive personalized medical insights and advice. For AI builders, this launch demonstrates how structured health data can power context-aware AI copilots, but it also highlights platform fragmentation and privacy considerations that affect adoption.

What happened

ChatGPT Health is now available on the web and the ChatGPT mobile apps for both Android and iOS. OpenAI is using its latest GPT-5.5 Instant and GPT-5.6 Sol models to handle medical queries. Users can ask for help comparing test results, analyze sleep and workout data, and get advice based on their health records. The feature also allows users to optionally grant ChatGPT access to their health data during normal conversations, so the AI can tailor recommendations for meal prep, restaurant choices, and activities.

OpenAI says it worked with physicians to extensively test the Health feature before release. However, the company faces ongoing legal scrutiny after a lawsuit alleged that ChatGPT provided dangerous medical recommendations that delayed treatment for a pulmonary embolism.

Why AI builders should care

ChatGPT Health is a concrete example of a personalized AI copilot that ingests sensitive structured data to improve response quality. For builders working on health-related AI products, the integration pattern is instructive: the system connects to existing health platforms (Apple Health, One Medical, Function Health) and hospital records, then uses model variants optimized for medical context. The GPT-5.5 Instant and GPT-5.6 Sol models are purpose-built for this domain, suggesting that model selection matters when handling health data.

The optional cross-context data access feature is also notable. If a user enables it, ChatGPT can reference lab results during a meal-planning conversation. This kind of persistent, context-aware personalization is a pattern many AI product teams are exploring, but it comes with heightened privacy and trust requirements.

Practical implications

For users, the immediate benefit is the ability to get AI-powered insights from their own health data without manually uploading files. The feature supports Apple Health integration and medical records from US hospital systems, making it easy to ask questions about trends in sleep, activity, and lab results. The optional data sharing across conversations means the AI can proactively consider health context, but users must opt in.

For developers, the rollout signals that OpenAI is investing in domain-specific model capabilities and health data partnerships. If you are building a health AI product, you should watch how OpenAI handles data residency, consent flows, and model safety in this feature. The privacy controls allow users to manage how their health data is used, which is a critical design pattern for any app handling sensitive information.

Caveats

The most notable gap is the lack of Google Health support. ChatGPT Health currently connects to Apple Health, One Medical, Function Health, and US hospital systems, but Google Health is not listed. OpenAI has not provided a timeline for Google Health integration, which means Android users who rely on Google Fit or other Google health services may not be able to connect their data.

Additionally, the feature is limited to US users aged 18 and older. International availability and support for other health data ecosystems remain unconfirmed. The ongoing lawsuit over dangerous medical recommendations also serves as a reminder that AI health advice carries real risks, and builders should implement robust safeguards and disclaimers.