Healthcare AI moved beyond the pilot phase in 2025.
From saving hundreds of lives to reducing costs by $20 million to more than $100 million annually, health systems started experiencing measurable payoffs from generative AI, ambient listening and predictive models over the past year.
Becker’s reached out to IT executives from leading health systems to ask for their most concrete AI returns on investment in the last 12 months:
Daniel Barchi. Senior Executive Vice President and CIO of CommonSpirit Health (Chicago): Of the 200-plus AI tools we now have deployed across CommonSpirit Health, one that provides a significant impact on patient outcomes is our AI-powered care gap closure system. It streamlined the way we identify and act on critical preventive care needs — things like cancer screenings and annual wellness visits.
By using AI to pull risk factors out of the EHR and calculate personalized screening timelines, and then using AI to automatically recommend the right orders to clinicians, we have been able to scale proactive care much more efficiently. In fiscal 2025 alone, we submitted over 61,000 care gap closure orders — a fivefold increase over the prior year — by automating the extremely complex adherence to cancer screening guidelines across breast, colorectal and lung cancer. It is good for our enterprise and even better for our patients and their ongoing clinical care.
Beyond operational ROI for AI, we also invested heavily in building the AI skills of our colleagues. We launched the CommonSpirit AI Learning Academy in May, offering nine long-form courses and 16 microlearning modules. The response has been significant, with more than 5,000 course registrations and 1,400 employees already upskilled. That skill-building has had a tangible impact on adoption, quality of use, and the speed at which clinical and operational teams bring AI into their workflows.
Lisa Stump. Executive Vice President and Chief Digital Information Officer of Mount Sinai Health System (New York City): The most significant ROI in 2025 came from an internally developed tool for malnutrition detection in inpatients. The malnutrition AI project identifies patients at risk for malnutrition and prioritizes them for assessment and intervention by the clinical nutrition team. Through early detection, intervention and documentation, it generated approximately $20 million in revenue impact.
Thomas Bentley, BSN, RN. Chief Information and Digital Transformation Officer of Ohio State University Wexner Medical Center (Columbus): Ambient listening technology, such as DAX Copilot, was the catalyst in reducing administrative burden of EHR documentation, allowing clinicians to focus fully on patient care and improving the overall experience for both providers and patients. Building on this foundation, AI-driven tools now summarize complex patient histories from massive EMRs in minutes, surfacing critical insights that were previously buried and reducing cognitive load. These innovations not only streamline care but also address patient safety by identifying significant results that might otherwise be overlooked. With annotations and references, the risk of hallucination is not only reduced but the clinicians are guided to the relevant information quickly.
Additionally, best clinical practices are being embedded directly into the EMR through actionable intelligence — flagging incidental findings and recommending evidence-based follow-up steps — helping clinicians avoid missed diagnoses and deliver safer, more efficient care. Adoption has been strong, and feedback from clinicians underscores the positive impact of these advancements on daily practice and patient outcomes. With next steps including draft shift and hospital summaries, our goal is to return the joy of medicine by getting clinicians out of the EHR and back to patient care.
Zafar Chaudry, MD. Senior Vice President and Chief AI, Digital and Information Officer of Seattle Children’s: From May to August, Seattle Children’s successfully piloted the ambient AI tool Abridge across 15 divisions involving 58 users, yielding overwhelmingly positive results. The 90-day pilot demonstrated significant clinical workflow benefits, with 76.7% of providers reporting a meaningful or positive impact on their daily routines by saving time and reducing the cognitive burden associated with charting.
Key usage metrics underscored the tool’s quality and effectiveness: Users gave an average star rating of 4.38/5, reported a 77% effort reduction, and Epic data confirmed a 15.5% reduction in time spent in notes per visit for the overall pilot population. With providers highly enthusiastic (90.7% recommending a full rollout), Seattle Children’s is now moving forward with systemwide implementation of this valuable technology.
Will Landry. Senior Vice President and CIO of FMOL Health (Baton Rouge, La.): Our most concrete return on investment has been ambient listening. For the providers who are actively using the technology, we’ve seen three areas of significant returns:
1. Reduction in total time our providers are spending in the EMR.
2. Reduction in time to close encounter notes.
3. Improved patient engagement in the visits where ambient listening was used.
Donna Roach. Chief Digital and Information Officer of University of Utah Health (Salt Lake City): In 2025, the University of Utah Health implemented and benefited from several forms of AI initiatives, including ambient listening, chart summarization, and early stroke detection in emergent cases. However, I believe our most meaningful investment in AI has been establishing dedicated AI labs and structured, clinical workflow testing environments.
While these initiatives are early and not yet producing direct financial ROI, they are delivering significant organizational value by creating safe, repeatable processes for evaluating tools, validating clinical use cases, and understanding integration requirements. This foundation allows us to make more informed decisions, avoid costly missteps, and accelerate responsible adoption. So the greatest return this year has been clarity — knowing what works, what doesn’t, and how to scale AI safely within a complex health system.
Scott Arnold. Executive Vice President and Chief Digital and Innovation Officer of Tampa (Fla.) General Hospital: Two highlights: The first is automated clinical documentation with Microsoft Dragon Copilot for physicians. Physician adoption has been quick to grow, and the unanticipated, positive side effect has been the unsolicited willingness for some providers to extend their careers with this technology. Many have suggested they were going to retire, but decided to stay longer given the advancement with documentation automation.
The second solid AI ROI for Tampa General Hospital has been our work with Palantir and using AI for our care coordination center. The team has created insights to improve operations and throughput (yielding a length-of-stay savings), as well as detect and monitor sepsis (saving over 700 lives).
Ryan Kenney. Vice President of Strategy Enablement at Nebraska Medicine (Omaha): Our most significant AI-enabled gains this past year were realized through applications focused on operational throughput and capacity management. Leveraging a people-process-technology approach, incorporating predictive analytics and machine learning models, we achieved and sustained a 5% reduction in average length of stay, effectively generating the equivalent of 37 additional inpatient beds at our academic medical center without physical expansion.
This incremental capacity enabled increased facility transfers and higher volumes of inpatient surgical cases, increasing margin for the health system. These downstream processes were also supported by AI-driven decision support and workflow optimization, reinforcing the system-level impact of these interventions.
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.
