The second day of Becker’s IT + RCM Conference brought together CIOs, CMIOs, chief revenue officers, legal experts, and clinical informatics leaders for a candid conversation about where health system technology is headed and what it will take to lead through the transition. The through-line across every session: the infrastructure era of health IT is giving way to something harder and more consequential. The organizations pulling ahead are not the ones with the most tools.
They are the ones with the clearest thinking about governance, culture, data, and the long game.
Here are 10 trends shaping that conversation.
1. EHR consolidation is the foundation while AI deployment is the real work
Health systems that spent years consolidating disparate EHR platforms are now discovering what that investment actually unlocks: the ability to deploy AI consistently, at scale, across an entire enterprise. In a fragmented environment, every AI tool required separate implementation across each platform, a cost and complexity that made meaningful scale nearly impossible. A unified foundation changes that equation entirely. The infrastructure question, for many systems, is now largely settled. What happens on top of that foundation, and how deliberately organizations build toward it, is where the real differentiation begins.
“Our goal is really for the technology to fade into the background, so our doctors and nurses can actually turn their attention to our patients and provide better care,” said Chris Carmody, CIO of UPMC.
2. Ambient AI has crossed the threshold from pilot to expectation
Ambient documentation has moved from a promising experiment to a standard of care at leading health systems — and the results are hard to argue with. Rigorous studies are showing meaningful reductions in clinician burnout, improvements in patient satisfaction, and time savings that compound across a physician’s day. The technology has largely sold itself wherever it has been deployed with care. The conversation at the frontier has moved past whether ambient works to what comes next: navigational agents, patient-journey AI, and tools that coordinate care rather than just capture it. For systems still evaluating ambient AI, the question is no longer whether to adopt but how quickly to do so responsibly.
3. AI governance is the most consequential capability gap in health IT
Across the industry, AI tools are being deployed faster than the processes to evaluate, monitor and course-correct them exist. The gap is not a technology problem; it’s a governance problem. Health systems that are getting AI right are building structures that look beyond the model itself to the workflow it will actually live in, taking a broad view of risk that includes patient autonomy, provider deskilling and equity implications and engaging frontline stakeholders — not just technical experts — in the evaluation process. The goal is not to slow adoption, but to build governance durable enough to outlast any single model or vendor. What is leading today will not be leading in 18 months.
4. Platform-first is winning — but health systems have a responsibility to keep innovation alive
The theoretical performance edge of best-in-class point solutions rarely survives contact with real-world adoption. A tool that is slightly better on paper but harder to use, harder to integrate, and harder to support will lose to a platform-native solution that clinicians and staff actually embrace. Health systems are increasingly orienting toward platform-first investment strategies as a result. But panelists were clear about the risk that comes with that orientation: the major EHR vendors are fast followers, not innovators. The startups and point solutions that health systems evaluate, pilot and champion today are what feed the platforms of tomorrow and systems that abandon that ecosystem entirely may find the pipeline of future innovation narrowing behind them.
“The right people doing the right process — then you get them on the right platform. That’s actually when you get things really unbelievable,” said Robert Bart, MD, CMIO of UPMC.
5. Bad data is the hidden ceiling on every AI initiative
EHRs are riddled with workarounds, configuration inconsistencies, and multi-entry-point gaps that create a false sense of data confidence. The smarter AI gets, the more those underlying data problems matter. Health systems that are making the most of their AI investments are not just deploying models; they are doing the hard, unglamorous work of building clean data foundations, closing interoperability gaps, and capturing information that has historically fallen outside the EHR entirely. Communication patterns, care coordination touchpoints, and behavioral signals that no structured field was ever collected.
“When we have more coherent and consistent data, then it can go to that action phase,” said Shahidul Mannan, Chief Data Officer at Boston Medical Center.
6. The CFO-RCM relationship is evolving into a strategic partnership
The traditional dynamic between finance and revenue cycle where the CFO focused on financial outcomes while revenue cycle focused on operational metrics is giving way to something more integrated and more consequential. Revenue cycle leaders are now embedded in site-of-service strategy, utilization management, medical policy disputes, and payer contract decisions that were once squarely in the finance lane. The most effective partnerships are built not on reporting structures but on shared metrics, shared language, and the ability to translate operational signals into financial implications before the CFO has to ask.
“You measure yourself not by how good you are at explaining your performance, but at influencing it,” said Beth Carlson, Chief Revenue Officer at Ohio State University Wexner Medical Center.
7. AI in revenue cycle is moving from task automation to strategic intelligence
The early wave of AI in revenue cycle was about automating discrete tasks like generating appeal letters, routing claims, flagging denials. The next wave is about intelligence: reading medical records against payer policy in real time, identifying systemic patterns in write-off data, and giving leaders the kind of analytical clarity that makes it possible to fix upstream problems rather than work downstream consequences. The near-term ROI case for AI in revenue cycle is strongest not where it replaces staff, but where it cleans the data flowing through the system reducing human variation, improving accuracy and making the metrics leadership relies on actually mean what they are supposed to mean.
“What I’m looking forward to the most is we’ll get cleaner and cleaner data if we reduce the variation of the human portion — so we know that our write-offs are more accurate as far as the bucket of why they happen,” said Gregory Wiles, Interim VP of Revenue Cycle at Atlantic Health.
8. Patient self-pay is rising and the collection strategy has to rise with it
The patient portion of healthcare revenue is growing, and the tools historically used to collect it are weakening. Medical debt no longer appears on credit reports. High-deductible plans are shifting more cost to patients who may not be prepared to pay it. Coverage disruptions, including sudden Medicaid and marketplace losses that health systems cannot anticipate, are creating waves of uninsured patients hitting facilities without warning. Point-of-service collection has become the dominant response, because the probability of collecting a balance after a patient leaves drops sharply. Health systems that have not modernized their financial counseling, pre-service estimation and front-end collection capabilities are increasingly exposed to a structural shift they cannot outrun.
“The patient portion has gone up significantly. There’s no question about it — and it’s getting harder and harder to collect,” said Mr. Wiles.
9. Personalization at scale is the next frontier for patient access
The ability to route the right patient to the right care at the right time, combining clinical acuity, operational capacity and individual patient preference, is emerging as one of the most valuable applications of AI in health systems. Getting access and navigation right up front has cascading effects downstream: fewer unnecessary ED visits, better care gap closure, higher patient satisfaction, and reduced burden on already-stretched clinical staff. Achieving it requires a level of data integration and workflow interoperability that most health systems are still building toward. But the organizations making the investment are finding that it changes the patient experience more fundamentally than almost any other intervention.
“Right care, right place, right time with the right clinician — if you get that right up front, you relieve a lot of the burden downstream,” said Khang Nguyen, MD, Medical Director of Care Transformation at the Southern California Permanente Medical Group.
10. Change management is still the hardest variable and the most underfunded
The technology is not the hard part of transformation. Getting thousands of clinicians, staff members, and administrators to change how they work and to trust that the change is worth it is the actual challenge. The organizations seeing results are not just deploying better tools. They are explaining the why at every level of the organization. They are using data transparently to demonstrate that new workflows are working. They are building time and space for staff to absorb change before the next initiative arrives. And they are building governance structures designed not for the AI model that is leading today, but for whatever model will be leading next year.
“Slow down to accelerate. Put all the right processes in place to help support the proper adoption and accountability and success that AI can bring,” said Chris Carmody, CIO of UPMC.
