As AI continues to emerge as a strategic focus for health systems, CIOs say AI capabilities embedded in EHRs are showing promise — but still have a long way to go.
A growing number of CIOs are helping guide their organizations’ AI strategies, with many seeing early benefits in areas like clinical documentation and operational efficiency. However, most say these tools remain in development and have yet to reach their full potential, according to a June 12 report from AI-based software provider Qventus and interviews with health IT leaders.
The report, which surveyed CIOs, CMIOs and other senior IT executives at medium and large health systems, found that the percentage of CIOs directing AI strategy jumped from 31% to 86% over the past year. But just 2% of respondents said their EHRs’ AI capabilities are fully developed.
That finding tracks with the perspective of Josh Glandorf, CIO of UC San Diego Health.
“Fully mature AI capabilities are still rare,” Mr. Glandorf told Becker’s. “There’s clear progress being made — especially in areas like ambient documentation, chart summarization, and operational efficiency. That said, we’re still early in the journey, and most solutions require further refinement and workflow integration to realize their full potential.”
Two-thirds of leaders surveyed said their AI strategies are still in development, while 20% described them as “limited or fragmented.” Despite that, more than half of health systems reported having a formal AI governance committee overseeing the development and deployment of AI tools.
Many leaders are prioritizing AI’s ability to support financial and operational goals. In the Qventus report, 54% cited improving operational efficiency and reducing costs — particularly in scheduling, staffing, patient flow and referrals — as their top strategic objective for care operations. When asked how they measure the return on AI investments, improved margins (26%) and cost reductions (24%) led the list, followed by staff productivity and clinician satisfaction (16% each).
Even as systems begin to see value from AI, many CIOs say today’s tools remain largely supportive rather than transformative.
“The biggest opportunities lie in advancing toward more agentic AI — tools that don’t just surface insights, but can initiate actions, follow up autonomously, and adapt in real time to evolving conditions,” Mr. Glandorf said. “Current systems are still largely reactive and require significant human intervention.”
Still, progress is being made. Mr. Glandorf pointed to new tools designed to draft responses to patient messages and optimize throughput as examples of operational AI that are beginning to make an impact. On the clinical side, intelligent chart summarization and ambient documentation are helping reduce cognitive burden for providers.
“Many of these tools are beginning to meet — and in some cases exceed — expectations,” he said.
Michael Pfeffer, MD, chief information and digital officer of Palo Alto, Calif.-based Stanford Health Care, said expectations around EHRs and AI should reflect where the technology currently stands.
“AI-driven tools in healthcare are still at the early stages of development and deployment, and I would not expect EHRs — or any IT platform used in healthcare — to be at the fully developed stage,” Dr. Pfeffer told Becker’s. “What is exciting is the current focus on delivering AI-enabled capabilities that enhance how we care for patients and improve the user experience and capabilities of our clinicians.”
Dr. Pfeffer said he anticipates more “value-added uses” for AI in EHRs that simplify workflows and support clinical decisions as the technology continues to evolve.
As vendors continue to roll out new AI features and health systems refine how they use them, CIOs appear cautiously optimistic. While AI-enhanced EHRs may not yet meet every expectation, leaders say the foundation is being laid — and the potential is becoming clearer.
