Why AI deals in healthcare have grown faster than other areas of tech — and what VCs are paying close attention to

This rapid tech adoption is a long time coming for the healthcare industry, which is infamously difficult to innovate. As healthcare organizations face labor shortages and pressures on their profits, startups are rushing in to solve their problems with AI — and many organizations are finally eager to try out the tech.

Still, the broader healthcare investment environment remains depressed, at least compared to 2021's funding frenzy, said Raysa Bousleiman, vice president of VC relationship management in SVB's life sciences and healthcare segment and coauthor of the report.

"Survive to '25 is what we keep hearing," she said.

While healthcare AI startups might have it better, they're not out of the woods. The relatively nascent sector still faces plenty of challenges, from competition with Big Tech players and other startups to navigating tricky regulatory and payment environments.

Here's where the most money in healthcare AI is going — and what these startups are up against.

Low-hanging fruit

Investors, startups, and healthcare organizations seem most eager to use AI to relieve providers of administrative burdens, a key driver in the industry's labor shortage, and reduce costs.

The category includes AI-powered virtual assistants, like those used by Innovaccer; clinical documentation tech like that deployed by Abridge; and revenue cycle management software built by startups such as Cohere Health.

Healthcare startups using AI for administrative tasks have grabbed $6.6 billion since 2021 and made up 42% of healthcare AI deals in the first quarter of this year, according to SVB. The segment faces few regulatory hurdles, the firm's report notes, and can lead to a clear return on investment for health systems and payers.

Still, some investors worry that the technology underlying these solutions, such as transcription technology used for clinical documentation, will quickly become commoditized.

Galym Imanbayev, a partner at Lightspeed Venture Partners and an investor in Abridge, said startups will need to create full-service solutions for enterprises to beat out other companies with similar foundational tech and train their models on high-quality data through key health system or payer partnerships.

"These large healthcare enterprises don't want just a plug-in tool. They need a complete solution," he said.

Even if new AI solutions claim to save providers significant amounts of time, Bousleiman said the tech has to integrate with existing clinical workflows for providers to want to use it.

"If a doctor has to open up a new window and go back and forth, it might actually create more friction in their day-to-day," she said.

However, one advantage of an early-stage startup, as opposed to a large tech player, is the ability to work closely with healthcare organizations to tweak the startup's products and gain providers' trust, Bousleiman said.

Startups that take a bottom-up approach to develop their AI tech alongside their customers could see big rewards in provider adoption, she said.

Partnering or competing with Big Tech

As healthcare startups using AI pick up steam, Big Tech giants are also digging into the growing sector.

Multiple legacy tech companies are training their AI models on medical data, like Google, which has said it's fine-tuning Gemini for healthcare use and rolled out a group of generative AI models piloted by health systems in December.

In some cases, Big Tech companies are pushing healthcare startups forward. Alphabet has backed startups including revenue cycle management company Fathom and care coordination platform Viz.ai. Nvidia's venture capital arm, NVentures, invested in medical scribe startup Abridge as part of the company's $150 million Series C. Nvidia has notched a number of partnerships with other healthcare startups, too, including Hippocratic AI, which is developing AI healthcare agents using Nvidia's tech.

But tech giants may also threaten the healthcare startup landscape. Microsoft's Nuance, which offers AI-powered clinical documentation and competes with startups like Abridge and Ambience, says its solutions are used by 77% of US hospitals.

As companies choose to offer low-cost or even free alternatives to Nuance's solutions, Define Ventures founder and managing partner Lynne Chou O'Keefe said that Nuance might choose to drop its prices to maintain its foothold.

That could be a big problem for its startup competitors since it's already challenging to get health systems to ditch their existing solutions like Nuance's in favor of a startup's tech.

Navigating regulations and reimbursement

AI-powered clinical analysis and diagnostic tools, like decision support tech, have raked in $5.3 billion since 2021, according to SVB. Tempus, the category's clear leader, will test the public market's appetite for its tech with a planned IPO this year that could value the company at up to $6.1 billion.

Still, these tools face plenty of challenges in making it to market. Technologies used to help diagnose patients often require FDA approval, and getting that clearance can be a long and expensive process for startups. Plus, diagnostic tools will have to secure payer reimbursement after they're approved, which means convincing health plans that the tech is worth the cost, Bousleiman said.

"Even if they achieve reimbursement, they still have that risk of monetization and profitability that the whole diagnostics sector faces," she said.

Lightspeed's Imanbayev said the firm is now leaning away from AI-powered clinical decision support and diagnostics, especially as health systems and payers are more eager to adopt other forms of healthcare AI.

"There are all these other low-hanging fruits that don't touch upon the very highly regulatorily scrutinized — appropriately so — realm of clinical decision support, and I'd rather focus there," he said. "But I think that time will come."