When OpenAI executive Fidji Simo was hospitalized for a kidney stone last year, she used ChatGPT to check a prescribed antibiotic against her medical history, which flagged a potentially dangerous interaction. That experience helped shape ChatGPT Health, unveiled this week.
“The resident was relieved I spoke up, she told me she only has a few minutes per patient during rounds, and that health records aren’t organized in a way that makes it easy to see,” said Simo, OpenAI’s CEO of Applications.
230 Million Weekly Health Queries
ChatGPT Health formalizes what has already become one of the chatbot’s most common use cases. More than 230 million people globally ask health and wellness questions on ChatGPT every week, according to OpenAI. The new Health experience creates a dedicated environment within ChatGPT where users can securely connect their medical records and wellness apps.
“One of the most common and meaningful ways people use ChatGPT today is for health and wellness information. By working with trusted brands like b.well, we’re able to make the user experience even more useful and personalized,” said Ashley Alexander, VP of Health Products at OpenAI.
ChatGPT Health operates as an architecturally isolated environment with purpose-built encryption. Users can connect apps including Apple Health, MyFitnessPal, Function, and Weight Watchers. Health conversations maintain a separate memory system from standard ChatGPT, and OpenAI says it will not train its models on personal medical data by default.
Promo
b.well Partnership Powers Medical Record Access
To enable medical record connectivity, OpenAI has partnered with b.well Connected Health, which operates a health data network spanning 2.2 million providers and more than 320 health plans, labs, and other sources.
“We know patients are already using ChatGPT for health information and advice. Allowing patients to securely and privately integrate their own health data will make those interactions much more contextually relevant and accurate,” said Kristen Valdes, CEO and Founder of b.well.
Her company’s SDK transforms connected health data into AI-optimized inputs, supporting both structured clinical data and unstructured information like clinical notes and diagnostic interpretations.
“Today, AI is only able to work across a small part of the patient record. The b.well Health SDK for AI finally enables AI to work across the full patient record, spanning multiple EMRs and with complete access to both structured data and unstructured data,” said Imran Qureshi, Chief AI and Technology Officer at b.well.
Built with More Than 260 Physicians Over Two Years
Physician involvement in building ChatGPT Health distinguishes it from typical consumer health apps that rely primarily on engineering teams. OpenAI developed the product over approximately two years with input from more than 260 physicians practicing across 60 countries.
Clinicians provided more than 600,000 feedback instances across 30 focus areas, evaluating model outputs for safety, clarity, and alignment with evidence-based medicine.
Clinical input shaped HealthBench, an evaluation framework that according to OpenAI contains 5,000 realistic health conversations and 48,562 unique rubric criteria. Unlike traditional medical AI benchmarks that test performance on licensing exam questions, HealthBench evaluates responses using physician-written rubrics reflecting real-world clinical judgment.
Privacy Without HIPAA Compliance
ChatGPT Health is not HIPAA compliant, as consumer health applications fall outside the scope of healthcare privacy regulations that govern providers and insurers. OpenAI acknowledges this limitation but points to layered privacy protections, including architectural isolation of health data and the option to enable multi-factor authentication.
Privacy advocates may note that OpenAI has faced recent court battles over user privacy, including being forced by a federal court to hand over deleted user conversations as part of ongoing litigation. OpenAI has called such requirements a “privacy nightmare,” but the orders create potential tension with ChatGPT Health’s promises around data control and deletion.
Users maintain control over their data and can remove medical record access at any time. Health conversations are excluded from model training by default.
Competition in Consumer Health AI
OpenAI enters a market where major technology companies are staking out positions with different strategies. Google also partnered with b.well in October 2025, potentially setting the stage for similar health features in its Gemini chatbot. Microsoft has been developing healthcare AI capabilities through its Dragon Copilot product line, which focuses on clinical documentation and physician workflows rather than direct consumer interactions.
Strategic positioning varies: while Microsoft targets healthcare providers seeking to reduce administrative burden, OpenAI is positioning ChatGPT Health as a consumer tool that helps patients navigate their own care. Microsoft also partnered with Harvard Medical School in October 2025 to strengthen its health AI capabilities.
ChatGPT Health is not OpenAI’s first venture into health AI. The company previously launched Thrive AI Health in July 2024, a partnership with Arianna Huffington’s Thrive Global focused on chronic disease management through AI-powered behavioral coaching.
According to OpenAI data, 70 percent of health-related ChatGPT conversations occur outside traditional clinic hours, when patients cannot easily reach their providers.
Availability
ChatGPT Health is launching via waitlist, with plans to expand access on web and iOS in the coming weeks. Electronic health record integrations and some app connections are available in the United States only. At launch, the feature is not available in the European Economic Area, Switzerland, or the United Kingdom.
For Simo, whose own hospital experience revealed how fragmented medical records can endanger patients, the launch represents a step toward giving consumers the same visibility into their health data that she improvised during her own care.
“I’ve heard many stories like this from people who are using AI to help connect the dots in their healthcare system that really wasn’t built to see the full picture,” Simo said.
