ChatGPT Health offers “dedicated experience” for health data, questions

OpenAI, the company behind ChatGPT, has announced the launch of ChatGPT Health: a “dedicated experience” within its ecosystem that includes bulked up privacy and security so consumers can upload health data, ask questions, and manage their wellness.  

“Health is already one of the most common ways people use ChatGPT, with hundreds of millions of people asking health and wellness questions each week,” a blog post says, adding that more than 230 million people globally ask health and wellness questions on ChatGPT every week.  

“ChatGPT Health builds on the strong privacy, security, and data controls across ChatGPT with additional, layered protections designed specifically for health—including purpose-built encryption and isolation to keep health conversations protected and compartmentalized.” 

Connecting personal health records to ChatGPT

Health users can securely connect their own data to their profile to construct a holistic, personal health profile, which can then be used to answer questions or supplement decision-making with a more personalized approach.  

“You can now securely connect medical records and wellness apps—like Apple Health, Function, and MyFitnessPal—so ChatGPT can help you understand recent test results, prepare for appointments with your doctor, get advice on how to approach your diet and workout routine, or understand the tradeoffs of different insurance options based on your healthcare patterns,” OpenAI says. 

OpenAI has partnered with b.well, a consumer-focused health information integration solution provider, on the data integration effort, to enable access to trusted healthcare providers in the United States. 

Direct input from the clinical community

The team has also collaborated directly with physicians to develop the service.  

“Over two years, we’ve worked with more than 260 physicians who have practiced in 60 countries and dozens of specialties to understand what makes an answer to a health question helpful or potentially harmful—this group has now provided feedback on model outputs over 600,000 times across 30 areas of focus,” the announcement says. 

This is a reference to HealthBench, the company’s tool for evaluating the capabilities, safety, and utility of its AI models for providing accurate and meaningful answers to health-related questions. 

However, clinical accuracy and safety have been a perennial pain point for large language models (LLMs) such as ChatGPT. Dangerous hallucinations are still a common concern, and LLMs often provide unevenly accurate answers on specific health issues.  

That being said, the announcement notes that the service is not intended to replace medical care, and shouldn’t be used to diagnose or treat any condition.   

A strong emphasis on privacy and security

Much of the announcement focuses on privacy and security concerns, making it clear that OpenAI is taking these issues seriously. 

Conversations that occur within the Health space are stored separately from other interactions with ChatGPT, and any information held in the Health space will not be used to train ChatGPT’s foundation models, which may offer some reassurance to privacy-conscious users. 

“Conversations and files across ChatGPT are encrypted by default at rest and in transit as part of our core security architecture,” the post adds. “Due to the sensitive nature of health data, Health builds on this foundation with additional, layered protections—including purpose-built encryption and isolation—to keep health conversations protected and compartmentalized.” 

All connected apps will need to meet privacy and security requirements, including collecting on the minimum amount of data needed, and consumers will be able to access information about what types of data may be collected by third-party applications. 

What does it mean for healthcare providers and patients?

With hundreds of millions of users already asking personal health questions and bringing ChatGPT readouts to their clinicians, the launch of Health isn’t likely to dramatically change this increasingly common patient behavior. It may, however, make it safer, slightly easier, and somewhat more accurate for consumers who choose to engage with AI in this way. 

By connecting personal health information, users will be able to get answers that actually fit their specific situations, which could cut down on the dreaded “the internet says I’m dying” syndrome, and could help to improve informed shared decision-making in the long term. 

Creating stronger privacy and security protections is also generally a good thing, although healthcare staff members still need to be told early and often that uploading protected health information (PHI) to ChatGPT under any circumstances is a no-go. 

Overall, it could be a good way to codify and contain a practice that’s already commonplace among patients of all types, and aligns with popular principles of increased patient access to personal data and greater engagement in monitoring and managing their own health. 

Assuming it succeeds where many, many previous attempts to wrangle disparate health data have failed, this new effort could be a positive step toward empowering patients and simplifying the process of managing wellness in between visits to healthcare providers. 

Right now, users can sign up for a waitlist to gain access to the service, which will roll out in phases across the ChatGPT user base.  

Jennifer Bresnick is a journalist and freelance content creator with a decade of experience in the health IT industry.  Her work has focused on leveraging innovative technology tools to create value, improve health equity, and achieve the promises of the learning health system.  She can be reached at [email protected].