OpenAI launches HIPAA-compliant healthcare tools, APIs for hospitals 

Hot on the heels of announcing a more secure, dedicated space for consumer health queries within ChatGPT, OpenAI has unveiled a series of products designed for hospitals and health systems with HIPAA compliance in mind.  

The two main offerings, ChatGPT for Healthcare and the OpenAI API, aim to give OpenAI a more official foot in the door of the burgeoning healthcare artificial intelligence ecosystem and make it safer and more secure for clinicians to leverage ChatGPT’s capabilities in the administrative and clinical spaces.  

“Healthcare is under unprecedented strain. Demand is rising, clinicians are overwhelmed by administrative work, and critical medical knowledge is fragmented across countless sources,” the announcement says. “Advances in models have significantly improved⁠ AI’s ability to support real-world clinical and administrative work, like helping clinicians personalize care using the latest evidence. Yet many clinicians still have to rely on their own tools because their organizations aren’t adopting AI fast enough, often due to the constraints of regulated environments.” 

“OpenAI for Healthcare helps close that gap by giving organizations a secure, enterprise-grade foundation for AI—so teams can use the same tools to deliver better, more reliable care, while supporting HIPAA compliance.” 

ChatGPT for Healthcare seeks to unify and simplify AI-enabled workflows

Healthcare organizations have been going all-in on AI adoption, but many have run into the same problems as a decade ago during the first EHR explosion: there are simultaneously too many and not enough tools to meet demand across the entire enterprise. The result is fragmented workflows, shadow IT, and frustrated users just trying to accomplish tasks as quickly and accurately as possible.  

Plenty of companies have taken a stab at fixing the problem with all-in-one workflows, and OpenAI is the latest to throw its hat into the ring. 

“ChatGPT for Healthcare is built to support the careful, evidence-based reasoning required in real patient care, while reducing administrative burden so teams can spend more time with patients,” the company says. “Organizations can bring clinicians, administrators, and researchers into a secure workspace with the controls they need to deploy AI securely and at scale.” 

The sandbox includes a variety of features designed to help achieve these goals, including: 

• Healthcare-specific, clinically vetted models powered by GPT-5, built for clinical, research, and operational work 
• Transparent citations accompanying evidence retrieval tasks, drawing on peer-reviewed research, public health guidance, and clinical guidelines, all accompanied by detailed citations to make fact-checking easier 
• Alignment with institutional policies and care pathways via integrations with enterprise tools so responses can incorporate specific recommendations tied to organizational priorities and guidelines 
• Shared, reusable templates for common tasks, including drafting discharge summaries, clinical letters, and prior authorization requests 
• Role-based access management to support enhanced governance within shared workspaces 
• Multiple options for data control, privacy, and security, including a business associate agreement (BAA) with OpenAI to support HIPAA-compliant use of the workspace and its tools 

Content shared with ChatGPT for Health won’t be used to train models, the company stresses, doubling down on its promises on the consumer side to keep sensitive information separate from its training initiatives. 

OpenAI has already rolled out the platform to a handful of notable health systems, such as AdventHealth, Baylor Scott & White Health, Boston Children’s Hospital, Cedars-Sinai Medical Center, HCA Healthcare, Memorial Sloan Kettering Cancer Center, Stanford Medicine Children’s Health, and University of California, San Francisco (UCSF).  

However, the announcement makes no mention of direct EHR integration, which could be a limiting factor in its success. If users still need to copy and paste each patient’s data into ChatGPT for Healthcare every time they want insights into recommended pathways or differential diagnoses, they could be saving less time and effort than imagined.  

In addition, more and more EHR companies are integrating similar features directly into their solutions, which could render OpenAI’s options redundant. Gaining traction in a notoriously fragmented and fickle field might not be as easy as relying on the name recognition that ChatGPT can bring to the party. 

OpenAI API creates a scaffolding for customized AI capabilities

The company is also promoting its ChatGPT application programming interface (API) that allows healthcare organizations to build their own AI applications powered by ChatGPT.  

“Teams are using our APIs to build healthcare applications including patient chart summarization, care team coordination, and discharge workflows,” the announcement notes. “Companies like Abridge, Ambience, and EliseAI are building capabilities like ambient listening, automated clinical documentation, and appointment scheduling for clinicians and patients.” 

Organizations can apply for a BAA for API-related work, as well, to ensure a HIPAA-compliant connection between their infrastructure and ChatGPT’s models. 

The bottom line: a breath of fresh air or the same old, same old? 

It’s not surprising that OpenAI is going fast and hard into healthcare, especially as the industry has embraced AI tools so quickly and comprehensively.   

While nearly every startup in existence is now “AI enabled” to some degree, and the vast majority of EHR companies have retooled their platforms to bring AI into as much of the workflow as possible, there are plenty of familiar Big Tech names that haven’t yet fully committed this time around after some stinging failures in earlier attempts to fix what’s broken about healthcare. 

Google and Microsoft, for example, have both crashed and burned in the health record management arena in the past. And while they are currently both are deeply involved in healthcare, they are focused on their own niches in the industry, leaving frontline clinical decision support (CDS) and document generation pretty wide open for OpenAI to scoop up, especially since ChatGPT is already a powerhouse in these areas in the consumer world.  

The key for the company will be outmaneuvering (or cleverly partnering with) EHR companies that have the built-in advantage of being the repositories of patient data – and convincing clinicians that this time around, everything will be different. 

The real test will come as ChatGPT for Healthcare rolls out to more organizations with varying degrees of AI maturity. How these organizations deploy the available tools, and whether they develop the governance structures necessary to see success, will be the defining factor for how useful and impactful the platform can be in real-world clinical situations.  

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].