ChatGPT Health Raises Big Security, Safety Concerns

ChatGPT Health, announced on 19 January 2026, promises a dedicated experience that securely brings your health information together with ChatGPT’s capabilities, but raises significant safety and data security questions. According to OpenAI, ChatGPT Health features encryption by default at rest and in transit and offers data controls such as multifactor authentication, though exact protections for health data on a regulatory level were not fully detailed.

OpenAI partners with network health data firm b[.]well to enable access to medical records, with the network built on Fast Healthcare Interoperability Resources (FHIR) standard APIs and trusted healthcare exchange frameworks, and it says the b[.]well network maintains strict identity verification, consent management, data security, and auditability controls.

Third-party connections, including wellness apps and medical records, must meet OpenAI privacy and security requirements, with users able to disconnect an app at any time, after which the app loses access. Critics warn that data shared in health contexts can be hard to completely delete and that data control may be lost once shared, emphasising the need for stringent minimisation, purpose limitation and protection, particularly for jurisdictions like the UK where GDPR applies.

Overall, experts advise caution about trusting health data to consumer AI assistants, given potential privacy and regulatory gaps and the broader risks of model errors or hallucinations.