OpenAI Launches ChatGPT Health with Isolated, Encrypted Health Data Controls

OpenAI has launched ChatGPT Health, a sandboxed space that lets users talk about their health and optionally securely connect medical records and wellness apps to receive tailored responses. The feature is described as having its own memory and purpose-built encryption and isolation to keep health conversations protected and compartmentalised, with conversations in Health not used to train OpenAI’s foundation models.

It is rolling out for ChatGPT Free, Go, Plus, and Pro plans outside of the European Economic Area, Switzerland, and the UK. Apps can connect to users’ health data only with explicit permission, and all Health apps must meet OpenAI’s privacy and security requirements, undergoing additional security review. OpenAI notes that Health has been evaluated against clinical standards using HealthBench, a benchmark OpenAI revealed in May 2025 to assess safety, clarity, and escalation of care.

According to The Guardian, Google AI Overviews has faced scrutiny, and OpenAI’s Health aims to provide safer, more controlled health interactions rather than replace medical diagnosis or treatment.