As AI moves deeper into hospital operations, health system COOs and strategy leaders say their roles are shifting from gathering information and overseeing workflows to coaching teams, building governance and deciding where the technology belongs. Many are using AI to reduce administrative burden, surface operational problems in real time and free clinicians to spend more time with patients.
But the 10 executives below are also clear about where they are pumping the brakes. Several warn that AI can’t fix broken processes and could simply automate them, while others say clinical decision-making must stay in the hands of clinicians, with named owners, monitoring and strong oversight before any tool goes live.
Editor’s note: Responses have been lightly edited for clarity and length.
Question: How is AI changing your role as a COO or strategic leader, and where are you being most cautious about adoption?
Peter M. Fleischut, MD. Executive Vice President and Chief Growth and Transformation Officer of NewYork-Presbyterian (New York City): AI is shifting the role of healthcare leaders from technology adoption to workforce transformation. The biggest opportunity is creating an AI-enabled workforce where clinicians and staff can use these tools to reduce administrative burden, improve decision-making, and focus more of their time on patient care. At NewYork-Presbyterian, we view AI as a strategic capability, not a standalone technology. We’re moving deliberately to embed AI across clinical, operational, and administrative workflows while maintaining strong governance, education, and oversight. My greatest caution is implementing AI faster than an organization can responsibly absorb it. Long-term success depends on preparing people and processes as thoughtfully as the technology itself.
Tim Riddell. Executive Vice President and COO of Ochsner Health (Jefferson, La.): AI is helping leaders shift more attention from routine work to higher-value decisions, strategy and organizational performance. At the same time, it has reinforced an important lesson: technology does not fix broken processes. In fact, AI often highlights where standardization, governance and workflow redesign are needed. My greatest caution is adopting AI before those foundational elements are in place. The most successful implementations pair technology with clear processes, accountability and human oversight so that AI enhances operations rather than adds complexity.
Doug McGill. Vice President of Quality, Strategy and Operations of Emory Healthcare (Atlanta): AI has become a force multiplier for many aspects of my work, particularly in communication, planning, information synthesis, and problem solving. Beyond helping accelerate the development of presentations and executive communications, I increasingly use AI as a thought partner to test hypotheses, challenge assumptions, and explore alternative perspectives. It helps move ideas from concept to a more structured and actionable form, while also accelerating research into operating practices, benchmarks, and improvement strategies.
Christopher Kane. Senior Vice President and Chief Strategy Officer of Phoebe Health (Albany, Ga.): A critical element is an AI governance work group that evaluates opportunities to apply AI to our clinical and business operations. A multi-disciplinary team with diverse perspectives mitigates the risk that we are moving too fast or too slow. Our version of an HOV lane: multiple people in the vehicle, ensuring that Phoebe moves at the optimal safe speed.
Eric Deaton. COO of Ballad Health (Johnson City, Tenn.): Historically, operational leaders spent a great amount of time gathering information before making decisions. AI helps surface patterns, exceptions, bottlenecks, and predictions in real time.
This aligns with Ballad’s planned use of AI-powered analytics for staffing and patient flow optimization.
A great deal of management time has traditionally been spent reviewing retrospective reports. AI allows leaders to spend less time asking, “What problem are we trying to solve going forward?” It allows us to skate where the puck is going.
The strongest value is generating actionable recommendations from operational data.
Matt Walsh. Executive Vice President and COO of Rush University System for Health (Chicago): COOs often act as coaches for their teams. Accelerating the use of AI adoption has increased the significance of that role. Ensuring our leaders get exposure to the right training and expertise to develop competency in this space is top of mind for me. I use the analogy of giving our leaders new glasses or contact lenses that help them identify business and clinical functions that could be improved with AI.
What I am most cautious about is ensuring we don’t use AI to automate bad processes. Fundamentally, to maximize the value of AI, we need to be rethinking our processes completely.
Kasey Paulus. Executive Vice President and COO of WellSpan Health (York, Pa.): Artificial intelligence is changing my role from oversight of operations to helping shape how innovation is embedded across the organization. As a strategic leader, I’m spending more time focused on governance, workforce transformation and ensuring new technologies are aligned with our mission and long-term strategy. At WellSpan, we’re using AI to improve access, reduce administrative burden, strengthen patient safety and help our teams focus more time on human interactions that matter most.
At the same time, we’re cautious about adopting AI without clear clinical value, strong oversight and measurable outcomes. The technology is moving quickly, but trust, safety and transparency have to move just as fast. Every AI solution we deploy is evaluated through a governance structure designed to ensure responsible, ethical and patient-centered use.
Patrick Tuer. Executive Vice President and COO of Encompass Health (Birmingham, Ala.): AI is changing my role as a COO by helping us streamline operational processes while keeping our focus firmly on patient-centered care. Our approach is to integrate AI tools into existing workflows in ways that support clinicians, not replace their judgment. Clinical decision-making will always remain in the hands of trained healthcare professionals.
One of the most impactful areas has been reducing administrative and paperwork burdens for clinical staff. By automating routine tasks, we can give caregivers more time to focus on what matters most: direct patient care and meaningful patient interactions.
We are also leveraging AI-powered virtual assistants to conduct post-discharge follow-up calls. These tools allow us to gather and analyze patient feedback at scale, giving us better visibility into how patients are progressing once they leave our hospitals. For example, we can identify whether patients have picked up their medications, scheduled follow-up appointments or encountered barriers to their recovery. This enables our teams to intervene earlier and provide additional support when needed. Where we are most cautious is ensuring that AI remains a tool for augmentation rather than a substitute for clinical expertise. We are committed to using AI responsibly, with appropriate oversight, privacy protections and safeguards to ensure that technology enhances care quality and outcomes rather than making decisions on behalf of clinicians.
Carol Gomes. CEO and COO of Stony Brook University Hospital (Stony Brook, N.Y.): AI is changing my role by requiring leaders like myself to think carefully about where technology can add value and where human judgment needs to remain central. At Stony Brook Medicine, we are using AI to support areas such as earlier and more accurate disease detection and research, while also looking at ways it can reduce administrative burden. My biggest area of caution is clinical decision-making. These tools can provide useful information, but the responsibility for care must remain with the clinicians who understand the patient and the full context of that decision.
Robert Wiehe. Senior Vice President and COO of UC Health (Cincinnati): AI has changed how I evaluate proposals more than it’s changed the work itself. Every operational problem now shows up with a vendor attached to it, and most of them are single-point solutions. Each one looks good on its own. Enough of them stacked up is a support and integration problem we’d be handing ourselves three years from now. So, before we ask whether a tool works, we ask whether it belongs in our enterprise platform or bolted onto the side of it.
Our core EHR partner is where this gets hardest. They’re built into everything we do clinically, so our default is to use what they build. But they aren’t always first to a capability, and waiting has a cost when something could relieve a real constraint today. What’s changed is that operations and our digital health team make that call together now. We say out loud which one we’re doing: solving the problem now and accepting we’ll migrate later or waiting because the integrated version will be better and it’s close enough to be worth it. Those used to be two separate conversations, and we ended up with tools nobody could support and constraints nobody could relieve.
We’re most cautious about anything touching clinical decision-making without a named owner and a monitoring plan. If we can’t say who’s accountable for the output and how we’ll know when it drifts, we’re not ready to turn it on.
