Something has shifted in how health system technology leaders talk about artificial intelligence. The tone is becoming more cautionary and the warning is not about AI failing to deliver on its promise but instead that health systems are moving too fast in the wrong direction.
Chris Carmody, CIO of UPMC in Pittsburgh, put the problem in structural terms.
“The most dangerous trend in health IT is treating AI as a collection of projects rather than a strategic capability,” Mr. Carmody said. “While experimentation is important, it can be challenging to achieve meaningful organizational impact.”
His enterprise is taking a deliberate approach, building the governance, leadership and enterprise adoption infrastructure needed before scaling. The measure of success, in his framing, is not the number of AI initiatives a system can point to.
“Success won’t come from having the most AI projects,” Mr. Carmody said. “It will come from executing a clear strategy for how AI can improve the way healthcare is delivered.”
That framing resonated across the conversations. Ryan Kenney, vice president of strategy enablement at Nebraska Medicine in Omaha, described the same risk in operational terms: health systems are layering new technology onto old workflows and calling it transformation.
“AI and automation are advancing faster than most health systems are redesigning the work around them,” Mr. Kenney said. “If we simply add new technology to old workflows, we risk creating more complexity instead of more capacity.”
Shannon Becker, vice president of IT for clinical and revenue cycle systems at Sentara in Virginia Beach, Va., drew the sharpest line between automation and improvement.
“When we automate a broken process, we do not fix it; we allow the problem to move faster, scale further, and become harder to see,” Ms. Becker said.
The downstream effects are concrete. Clinicians become overwhelmed by tools meant to help them and patients fall through the gaps between disconnected teams. The organization loses its capacity to absorb meaningful innovation because each new capability must navigate years of accumulated variation.
“AI will amplify whatever foundation we give it,” Ms. Becker said. “If the foundation is fragmented, AI will accelerate fragmentation. If the foundation is thoughtful, standardized, and clinically grounded, it can genuinely transform healthcare.”
That transformation takes a completely different mindset and strategic approach to designing teams and workflows. Andie Romaniuk, senior director of information management applications at Ann & Robert H. Lurie Children’s Hospital of Chicago, framed the challenge as a question most organizations are not yet asking.
“Instead of asking, ‘How do we implement AI?’ we should be asking, ‘What work no longer needs to exist?’” Ms. Romaniuk said. “The organizations that will lead the future of healthcare won’t be those with the most technology. They will be the ones that use technology to empower teams, reduce friction, create capacity, and improve outcomes for patients, families, caregivers and the organization alike.”
Even the roll-out messaging needs an overhaul. Health systems have become familiar with bringing on new technologies, doing kick-offs and measuring utilization. But that’s not enough to push past the technology and achieve the best results.
“Transformation doesn’t happen when technology goes live,” Ms. Romaniuk said. “It happens when people see themselves in the solution, embrace new ways of working, and experience the value firsthand.”
Hetal Rupani, senior director of business intelligence and analytics at Johns Hopkins Medicine in Baltimore, pointed to a second structural blind spot: the unclear division of responsibility between EHR vendors and health systems.
“EHR platforms are foundational, but the responsibility for making technology scalable, usable, and clinically meaningful sits with health systems,” Ms. Rupani said. “When that distinction is not clear, organizations can end up with expensive systems that are technically live but operationally underperforming.”
Andrew Cooper, chief digital and access officer at NCH in Naples, Fla., offered the most direct formulation of the core tension. Innovation and complexity can look identical from a distance.
“The most dangerous trend is complexity disguised as innovation,” Mr. Cooper said. “We’re going to see an explosion of AI and digital tools, but if we’re not disciplined about how we deploy them, we’ll simply create more noise for clinicians and patients.”
The path forward, across all six perspectives, requires sequencing work that health systems have often skipped in the push to show AI momentum: simplifying before automating, redesigning workflows before deploying tools, and measuring results in clinical and operational terms rather than implementation counts.
“The goal isn’t to implement more technology,” Ms. Romaniuk said. “The goal is to create more time for care.”
At the Becker's 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, taking place September 14–17 in Chicago, healthcare executives and digital leaders from across the country will come together to explore how AI, interoperability, cybersecurity, and revenue cycle innovation are transforming care delivery, strengthening financial performance, and driving the next era of digital health. Apply for complimentary registration now.
