Why hospitals and tech companies are teaming up so much

Lately, hospitals and tech companies have been speaking the same language—or should we say LLM. They have been announcing seemingly nonstop partnerships to create new AI tools for healthcare.

The announcements are part of a push by tech companies to gain access to valuable health data. For hospitals, the partnerships offer AI expertise and the opportunity to implement AI tools faster.

Let’s just name a few: Mayo Clinic and Microsoft, Phillips and WellSpan Health, Samsung and Mass General Brigham, Boston Children’s and OpenAI, Cleveland Clinic and Palantir, HCA Healthcare and Google, and Amazon One Medical and Rush University System for Health.

Provider and tech partnerships are not necessarily new. They were very popular during the Covid-19 pandemic, for example, amid the rise of telehealth. Take Best Buy and Atrium Health for example, or Microsoft and Epic.

But now with the rise of generative AI in healthcare, from the use of scribes, to research tools and back office systems, tech is becoming a bigger part of the medical industry, experts told us.

“It is definitely accelerating,” Bill Fera, principal at Deloitte said.

So why is this happening? And could big tech leave a long term mark on healthcare?

Pros for tech companies. The reason tech companies want to work with hospitals, Fera said, is because it gives them access to the medical data that is necessary for building quality products.

“We’re finding that general data and general models don’t translate as well in healthcare,” Fera said. “So there’s a need to really work on real data and real situations to create high fidelity generative AI models and agentic models.”

For example, sepsis models, which predict a patient’s risk for developing the condition, work better when the models are trained specifically on local hospital system data, Fera said.

A hospital’s patient data is also regulated under HIPAA, so partnering with hospitals gives tech companies more access than they usually would.

Further, people aren’t generally as trusting of big tech as they are of healthcare, Andrew Rebhan, digital health advocate and market researcher for Vizient Strategy Intelligence, added.

“The tech companies themselves are saying, ‘look, we need to rely on trusted physician and clinician input and expertise,’” he said, to build trust with tech.

Last, these partnerships give tech companies a chance to have their products endorsed by the medical industry.

“It’s a way for them to say, ‘Look, we’ve actually got the brand A, brand B, brand C of these large health systems that are willing to work with us and vouch with us to test these products out,” Rebhan said.

Pros for hospitals. For hospitals, partnerships present new ways to implement tech that could help ease administrative and clinical work.

“Hospital systems are also looking to ease the burden on their providers and their clinicians, and so [AI] is a way to do that,” Fera said.

Working with outside companies gives clinical staff a deeper understanding of technology. It also provides an opportunity for hospitals to develop the tools and influence how they are made, Rebhan said.

“Partnerships are really just a means of quicker access, speed to market value, and then there’s the third angle to it,” Rebhan said, that many big health systems have innovation arms and can invest in the companies for potential ROI.

Mayo Clinic, for example, is co-developing a “frontier AI model” with Microsoft and will own the product in the end.

“We’re in an extremely margin constrained industry, and so anything that you can do to help from that perspective is welcome,” Fera said.

What then? As the number of partnerships expands, big tech companies are also releasing their own healthcare tools. Open AI has ChatGPT Health, which allows consumers to upload their medical records, and Amazon has One Medical, which offers in-person and telehealth primary care.

“If they’re building their own separate enterprise solutions or direct-to-consumer solutions, how does that ultimately affect the influence and historical control of patient access that health systems have had?” Rebhan said.