AI-driven uncertainty is reshaping healthcare IT (HCIT) dealmaking as buyers tread with caution and rethink valuations across the sector.

While the sector remains fundamentally healthy, dealmakers say AI is increasingly impacting pricing, diligence and transaction structures. Against that backdrop, some exits are being postponed, amid an increasing focus on how much of today’s functionality could ultimately be dislodged by AI.

Year-to-date, there have been 36 healthcare IT transactions in North America involving both strategic acquirers and financial sponsors, a roughly 30% decline from the same period last year and the lowest level since 2023, according to Mergermarket data.

Several health-tech assets expected to come to market this year have already been pushed into 2027 as sponsors reassess where multiples may settle, according to market participants.

Large HCIT deals remain mostly on pause as buyers recalibrate around AI’s impact, said two sector bankers. The exception is the ongoing sale of Warburg Pincus- and Berkshire-backed Ensemble Health, which had progressed to the second round as of mid-April and is being marketed on more than USD 700m in EBITDA, as previously reported by this news service.

Source: Mergermarket, data correct as at 7-May-26

Universal theme

Dealmakers say AI is now embedded in nearly every HCIT process. Dudley Baker and Luiz Greca, managing directors of healthcare technology at Houlihan Lokey, said AI enablement has become a universal theme in their transactions, with four of the last five deals they advised on including AI components.

Every HCIT company is expected to present a credible AI strategy, Baker said, while Greca added that the absence of one is increasingly viewed as a red flag. AI’s ability to streamline workflows and reduce staffing needs has also become central to how sellers position value.

Baker pointed to recent transactions such as Apploi, VisiQuate, DeepIntent and Forcura/Medalogix combination, all of which embedded AI in their solutions, as examples of how AI capabilities are increasing in deal narratives.

Deal strategy is shifting accordingly. After a period dominated by platform investments, PE sponsors are expected to increasingly pursue smaller bolt-on acquisitions of AI and agentic AI businesses to future-proof existing assets, another sector banker said.

Reset

Meanwhile, sellers are being forced to rethink exit strategies, with some turning to alternatives such as continuation vehicles (CVs). NewSpring, for example, pivoted to a CV transaction for Verisma after bids fell short of valuation expectations, with sources attributing the gap to AI-driven market fallout, this news service reported in April.

More broadly, tech-adjacent HCIT businesses — particularly those with workflow automation exposure — are seeing multiples compress as buyers reassess long-term defensibility.

Assets that once commanded around 18x EBITDA are in some cases being re-rated closer to 14x, said another sector banker. For sponsors that bought at peak pricing, shifts of that magnitude make exits harder without significant growth, extending hold periods and delaying launches.

A potential test case is WellSky, backed by TPG Capital and Leonard Green & Partners. The company is expected to be marketed on roughly USD 400m of EBITDA, with a target valuation of around 20x, though some sources said that may prove difficult in the current environment, this news service reported in March.

Valuation pressure is not uniform, however. In contrast to broader software, where AI’s impact is more fully priced in, HCIT is seeing a more uneven, sub-sector-specific effect, one banker said. Some segments are expected to prove more resilient than others over the next 6–12 months.

The premium for AI-enabled HCIT assets remains real but increasingly selective, said Bob Farrell, CEO of mPulse. Buyers are paying up only for companies with proprietary models, defensible data, strong customer footprints and measurable outcomes such as HEDIS or STAR improvements. Multiples fall quickly where diligence uncovers AI washing, reliance on a single foundation model, unclear data rights or weak outcomes. Privacy, bias and regulatory gaps are also weighing on valuations for companies without mature AI governance.

Reality check

Others point to lingering uncertainty, as investors work to assess how AI is impacting underlying economics.

The disconnect between fundamentals and valuation is widening across enterprise, said Chris Dorn, managing director of healthcare technology at Fifth Third Securities.

While the cost of generating new software features is approaching zero, the ongoing maintenance burden of AI-enabled tools is often underestimated.

“You aren’t paying for software, but you are hiring additional people to maintain newly developed internal tools. Is the net cost to the customer unchanged?” Dorn said.

This uncertainty is feeding directly into deal dynamics. Dorn said many PE investors are simultaneously too cautious on legacy software and too optimistic on AI-native models, widening bid–ask spreads and derailing processes. Investment committees increasingly view AI disruption as a real risk, even for deeply embedded platforms.

At the same time, not all assets are equally exposed. Healthcare software companies with differentiated data are better positioned to benefit, Dorn said, pointing to R1 RCM’s creation of R37, an internal AI R&D lab, as an example of how cash‑flowing HCIT businesses can build AI-native solutions organically.

There is also growing interest in services businesses where buyers see scope for automation-driven efficiencies, though Dorn cautioned that the opportunity is widely recognized and may not remain undervalued.

Despite the focus on AI, operators emphasize that it is not yet the primary driver of transactions. Buyers are still underwriting on recurring revenue, contract density, and servicing economics, with AI viewed as post‑close optimization engine than a standalone acquisition driver, said Saul Mateos, CFO of Gain, a provider of AI-enabled medical lien management solutions.

“If it still works without the AI layer, just more slowly, it’s an enhancement, not differentiation,” Mateos added, explaining the diligence bar he uses to assess whether a product is truly dependent on AI. True defensibility emerges when workflows break without AI and when underlying data and workflow depth cannot be replicated by simply wrapping GPT or Claude, he added.

Recent deal patterns reflect this divide. Strategics are paying up for AI-native teams, often using earnouts around retention, while PE buyers remain focused on platforms where automation can expand margins, Mateos said.

Split market

Sellers with strong AI narratives or embedded data assets are outperforming, while those without a clear strategy are facing steeper valuation haircuts, bankers said. Sponsors are also shifting away from pure platform creation toward acquiring AI capabilities, reflecting a broader rethink of how value will be created and protected.

The result is a more divided market: assets with credible AI upside are commanding attention, while those exposed to automation risk are struggling to clear valuation hurdles.

Looking ahead, the most attractive assets will combine proprietary data, workflow depth and automation-driven unit economics, along with management teams that can articulate inference costs at a per-transaction level — a new competency separating durable AI businesses from decorative ones, Mateos said.