ChatGPT Health: OpenAI’s New Platform

ChatGPT Health: OpenAI’s New Platform That Connects Your Health Data and Apps for Personalized, Protected Responses

This solution allows users to integrate their medical records and applications such as Apple Health or MyFitnessPal, exploring habits and analyses under strict privacy controls and advanced personal data protection options. OpenAI has introduced ChatGPT Health, a platform that integrates artificial intelligence with personal health data to provide personalized, confidential responses to questions about wellness and medicine. The launch responds to the strong demand for health information, as more than 230 million people worldwide request health and wellness advice from ChatGPT every week, according to the company.

The main goal is to enable users to securely connect their medical records and health apps, such as Apple Health and MyFitnessPal, to better understand their analyses, habits, and routines — and to prepare in an informed way before medical appointments or resolve common questions about diet and exercise.

Although ChatGPT Health brings innovations in privacy and data control, OpenAI itself stresses certain fundamental limits. Project leaders note that the platform is not designed to perform diagnoses or provide medical treatments.

In fact, there is an explicit emphasis on encouraging users to always consult healthcare professionals for relevant symptoms or critical situations. Multiple safeguards have been deployed to avoid alarmist responses and to automatically redirect users to in-person resources when the system detects signs of danger, especially in sensitive areas such as mental health.

In developing ChatGPT Health, OpenAI engaged more than 260 physicians from 60 countries across dozens of specialties. Over two years, this expert panel evaluated over 600,000 model interactions, helping establish rigorous clinical criteria prioritizing safety, clarity in communication, and relevance of recommendations.

The system incorporates its own evaluation standards, such as the HealthBench framework, which focuses on measuring the practical usefulness of responses and their alignment with real medical judgment, rather than purely theoretical accuracy.

In practice, ChatGPT Health allows users to link apps such as Function and MyFitnessPal to analyze eating and physical habits, or services such as b.well, a U.S. network for secure medical data exchange used by more than 2.2 million healthcare providers.

This connectivity is optional; according to OpenAI, external applications can access information only with the user’s explicit consent. The company enforces a minimal data collection policy and conducts specific security reviews before integrating any new platform into the Health environment.

Security and control: a separate architecture for confidentiality

The ChatGPT Health space operates in isolation: medical conversations and files are stored separately from other ChatGPT interactions and have a dedicated memory to prevent data leaks.

All conversations are encrypted both in transit and at rest, thanks to an architecture specifically designed for sensitive data. ChatGPT also recommends enabling multi-factor authentication (MFA) to enhance protection against unauthorized access.

Users can individually review or delete their stored medical histories within the system and manage the access permissions of connected apps via the settings panel.

It’s worth noting that data hosted in ChatGPT Health is not used to train OpenAI’s language model, reducing risks linked to secondary use of personal information.

Legal issues, availability, and geographic scope

Despite the announced data protection measures, the environment does not implement end-to-end encryption, and data may have to be handed over to authorities in response to court orders or emergencies, according to OpenAI’s head of digital health.

Additionally, the company confirmed that compliance with the U.S. HIPAA privacy law does not apply in consumer settings, only in clinical or professional contexts.

For now, the feature allowing medical record and app integration is limited to the United States, while general product access is being rolled out gradually to selected users of both free and paid versions of ChatGPT. The service currently excludes the European Economic Area, Switzerland, and the United Kingdom.

To access it, users must join a waitlist, and the company plans to expand ChatGPT Health’s availability to more regions in the coming weeks.

Risks and controversies: from anxiety to mental health

During the presentation conference, OpenAI executives acknowledged risks associated with misuse of ChatGPT Health, particularly in sensitive areas such as mental health and the potential worsening of anxiety or hypochondria.

Recent cases and reports have documented erroneous or harmful responses from AI systems in health contexts, including potentially dangerous guidance on diets, medical tests, or risky behaviors.

The company says it has adjusted its model to avoid alarmism and to prioritize redirection toward conventional healthcare services in situations of risk.

However, OpenAI’s cybersecurity record is not flawless: in March 2023, a data breach exposed some users’ names, emails, and partial payment details — underscoring the importance of using the system cautiously, especially when handling sensitive information.

Links

• Introducing ChatGPT Health – OpenAI
• Sell pre-owned AI device online – iGotOffer

ChatGPT Health: Can You Really Trust AI With Your Medical Advice? | The Big Question [Video]

Video uploaded by NDTV Profit on January 8, 2026.