AI is reshaping the software business model across the economy, and healthcare software is right to ask whether it is next. Our view is that the threat is real but contained: AI accelerates development for everyone and has become an essential tool, but the barriers that protect healthcare software lie outside software development, and AI leaves them largely intact.
Across the wider economy, AI is reshaping the software business model: it has lowered the cost of building software, unsettled per-seat pricing, and opened the door to a new generation of AI-native competitors. Healthcare IT runs on the same technology, which raises a natural question for investors weighing the durability of these businesses: is the sector exposed in the same way? In practice, the question resolves into three more specific ones: could customers now build their own tools, could the large EMR incumbents pull more development in-house, and could AI-native start-ups emerge as serious competitors? We take each in turn below.
Could customers build their own software in-house?
Probably not at any meaningful scale. The real cost of a clinical-grade product has always lain less in writing it than in regulatory approval, validation, and ongoing maintenance, and that is exactly the part AI does not lighten. Healthcare providers rarely keep the dedicated development, validation, and maintenance teams such a product demands, and even where coding skill exists, it seldom stretches to the ongoing burden of running a regulated system. The MDR in-house exemption is demanding in its own right: a provider must meet full safety, performance, and quality requirements, and show that no equivalent CE-marked product is already on the market. AI may shorten the build, but it does little to ease the regulatory and operational load that follows.
Could EMR incumbents pull development in-house?
More likely in part than wholesale. For now, the large EMR players are absorbed in embedding AI into their core platforms, competing on interoperability, and adjusting to new pricing models, leaving limited bandwidth for the many fragmented add-on categories around them. They are unlikely to take these on unless sustained customer dissatisfaction makes the investment worthwhile. In many cases the more valuable path runs the other way: keeping interfaces open and integrating best-in-class third-party solutions, rather than rebuilding every function in-house. Selective absorption, not wholesale insourcing, looks like a more realistic pattern.
Could AI-native start-ups become serious competitors?
Only gradually, if likely at all. Faster development does not necessarily mean faster market entry. A new entrant still must secure regulatory conformity, achieve EMR integration and interoperability certification, build clinical and use-case references, and work through public procurement cycles that stay multi-year however quickly the software was written, and AI shortens almost none of this. Established providers, meanwhile, tend to be deeply embedded: mapped into EMR data structures, woven into clinical workflows, and held under multi-year contracts. Switching costs can stay high even when a newer alternative looks competitive on its merits.
Where does AI change the game?
None of this means AI is unimportant in healthcare IT, only that its role is different from the one the broader SaaS narrative implies. AI is compressing development cycles across the sector, raising the pace at which credible providers must ship new features and respond to customer needs. Buyer expectations are rising with it: hospitals and other clients increasingly expect AI-enabled functionality as standard and tolerate fewer manual workflows than before. Competition, while constrained at the structural level, sharpens at the product level: incumbents and entrants alike can move faster, iterate more, replicate winning solutions, and demonstrate value more visibly. AI may not decide who wins the market, but it raises the bar for staying in it. Beyond pace, AI also changes how some forms of competitive advantage are built. Incumbents with years of real clinical deployment data can train proprietary models that newer entrants cannot easily replicate, turning an existing position into a renewed one. Building that kind of AI capability is demanding work in its own right, and the window to do so is open now, but will not stay open indefinitely.
The bottom line
The three concerns point in the same direction. AI is a powerful accelerator that no serious provider can afford to ignore, but in healthcare IT it is unlikely to act as a competitive equalizer. What tends to decide who wins (i.e., regulatory expertise, workflow embeddedness, and trusted clinical data) sits largely outside software development, and AI leaves it mostly intact. For investors, that makes healthcare software more resilient to AI disruption than the broader SaaS narrative would suggest. The risk worth watching is less that new competition arrives, and more that incumbents treat AI as optional and fall behind on the pace and product evolution it now demands. The providers worth backing are those that hold these structural strengths and take AI seriously enough to defend and build on them.
Get in touch with our Strategy & transactions team
Read more From our case studies
Western Uusimaa wellbeing region sought international knowledge to refine its personal doctor model
Western Uusimaa wellbeing region sought international knowledge to refine its personal doctor model The Western Uusimaa wellbeing services county (LUVN) is developing a personal doctor model…
Read more
Helsinki University Hospital sought clarity on how the EHDS Regulation will impact patient information systems
Helsinki University Hospital (HUS) sought clarity on how the EHDS Regulation will impact patient information systems The European Health Data Space (EHDS) is set to transform the use of…
Read more
North Savo Wellbeing services county enhances elderly services – “Staff satisfaction has increased as well”
Data-driven management has brought greater consistency to services and enabled resources to be allocated where they are most needed. Wellbeing services county of North Savo has significantly…
Read more