Software vs. services: What AI is forcing healthcare to rethink
The discussions surrounding artificial intelligence often present as a clean break from the past. They are not. The recent expansion of AI into healthcare has forced a choice that has faced providers since the early days of EHRs — a choice that is changing what buyers expect technology to do, and what they are willing to pay for.
The choice is simple. Providers can choose between purchasing software products — seats, licenses, and updates — or they can choose to partner with services that deliver an outcome.
The recent development of AI offerings for providers has accentuated this long-standing choice and expanded where in workflows it can occur. Buyers are no longer evaluating technology solely on features and functions, but on outcomes — what work gets done, how reliably, and with what measurable impact. AI didn’t invent this expectation, but it has made it unavoidable. We used to call technology products that deliver service outcomes “software-enabled services.” This is what has defined our business since its inception. The AI shift has evolved this into “AI-enabled services.”
This new era raises a different set of questions for healthcare leaders. How should they evaluate technology in this new environment? What does partnership really mean when software is expected to deliver services directly? And how does this shift intersect with the broader pressures facing healthcare over the next decade? Do AI solutions offer a complete service across all use cases, or do they need to be backed by staff for edge cases? Can these services offer resiliency when connections go down?
The reality is that software companies have always been services companies in disguise. AI simply makes that reality explicit.
Why outcomes are the real product
For years, much of healthcare technology has been sold through licensing models that emphasize access to tools rather than responsibility for results. Features mattered. Roadmaps mattered. But the burden of making those tools deliver value often fell on the customer.
What AI is doing is changing that equation. When an AI agent can automate work directly — whether that’s administrative tasks, coordination, or decision support — the line between product and service starts to blur. The reality is that software companies have always been services companies in disguise. AI simply makes that reality explicit.
This shift puts new expectations on technology providers. A services orientation means being accountable for outcomes, not just functionality. It means thinking in terms of measurable efficacy and operational savings, not just capital expenditure and licensing. It means delivering customer success, not just support. And it means adopting a partnership mindset, where success is shared and performance is visible. Features and functions still matter. They’re just not sufficient.
What AI changes in healthcare — and what it doesn’t
In healthcare, productivity gains have historically been difficult to achieve. Demand continues to rise. Staffing shortages persist. Administrative work consumes an enormous share of resources — by some estimates, roughly 30% of healthcare spending is tied up in administrative tasks. AI creates an opportunity to scale the “backstage” of healthcare: the workflows, preparation, coordination, and follow-up that enable care to happen and get paid for. That work doesn’t disappear, but it can be delivered more efficiently and at a lower cost through automation.
What AI doesn’t do is replace the fundamentally human aspects of care. Medicine is, after all, an art as much as it is a science. Empathy, trust, and connection still matter. The challenge is determining where AI-driven efficiency helps — and where human interaction remains essential. That balance will shape how healthcare organizations deploy these technologies over time.
Why ecosystems matter more than ever
Another implication of this shift is that AI agents are only effective within the context provided to them. Agents without full patient context, or without full connectivity to transact, are limited and, worse, risk errors. As such, ecosystem access is becoming increasingly important. By the same token, different AI models and their downstream agents are optimized for specific tasks — so the ability to connect multiple agents into a cohesive service or outcome is also critical.
From a buyer’s perspective, the question is no longer just whether a product works, but whether the surrounding system enables its success, and whether it further provides the customer the ability to adapt and evolve with change. If anything, the last 24 months have shown us how fast assumptions can change, and how easily purchasing decisions can be regretted. Key questions for providers include: How does the choice I make preserve flexibility? Will it evolve and provide me access to innovation as it unfolds? Does the technology support integration of these products natively? Does it allow new capabilities to emerge without rebuilding everything from scratch? Or does it constrain innovation by design?
We’re already seeing AI partners converge from point solutions into broader platforms. That trend reinforces the need for open, dynamic ecosystems where services can work together. In a services-oriented world, the ability to collaborate matters as much as the capabilities themselves.
What this means for care delivery
All of this is unfolding against a challenging backdrop. Healthcare is facing a cresting wave of demand, persistent workforce shortages, and growing pressure on access. The system has limited room to take on more administrative burden — and clinicians are already paying the price, with downstream effects for patients.
In that context, expecting technology to deliver outcomes isn’t just a preference — it’s a necessity. Reducing administrative cost through automation, allocating work more effectively across care teams, and being deliberate about how physician time is used are essential steps toward sustainability.
At the same time, care itself remains a deeply human experience. While AI can support interactions, most people want to feel cared for in their most vulnerable moments, not processed. The opportunity is to use technology to scale what happens behind the scenes, improve care decisions, and create focus so that the moments that matter most can become more human, not less.
The promise of this moment is significant — AI-powered services, connected to the healthcare ecosystem, can unlock real shifts in care — physicians gaining more choice in their practice, supported by services that travel with them rather than institutions that constrain them. Care could be delivered sustainably in more settings, in more flexible ways, without sacrificing quality or connection.
That’s not a guaranteed outcome. But it’s a direction worth aiming toward.
AI is forcing healthcare to confront long-standing questions about what software is for and what providers should expect — and whether there is a better way to deliver care. The answers won’t arrive all at once. But the organizations that engage these questions seriously — with a focus on outcomes, partnership, and care delivery — will be the ones best equipped to sustain access, support clinicians, and navigate what comes next.
