In 2026, health system IT leaders are focused on scaling artificial intelligence, moving from pilots and proofs-of-concepts to operationalizing AI, but they are facing major barriers in these efforts.
Seventy-four percent of health system leaders cite electronic health record (EHR) vendor reliance as an obstacle to executing their AI strategy, new research from Qventus found.
More than 60 chief information officers, chief AI officers, chief medical information officers and other senior IT leadership members at health systems across the country were surveyed for the April 9 report. Leaders at these organizations, including HonorHealth and Rochester Regional Health, were asked how they are implementing AI tools and navigating EHR dependency.
“I think that the cost of waiting for Epic or Oracle, or any of them, is you might lose out,” said Matthew Anderson, M.D., CMIO at HonorHealth, in the report. “There’s a late-mover disadvantage. Not a first-mover advantage, but there is a late-mover disadvantage.”
However, the report found that surveyed organizations are less willing to hold out for EHR systems to develop AI solutions, as only 22% of respondents said they would wait for an EHR feature, down from 52% in 2025.
When asked whether health systems would wait 18 months for an AI feature from an EHR versus deploying a third-party vendor solution in three months with guaranteed return on investment, 40% would move to a third-party solution now, 38% say it depends, and only 22% would wait, the survey found.
AI tools have become commonplace in healthcare organizations across the nation. A March survey from Eliciting Insights found 75% of U.S. health systems are using at least one AI platform, up from 59% in 2025. Moreover, 50% of respondents indicated their systems use three or more AI applications.
Alongside increased use of and spending on such tools comes increased pressure to implement them, as 65% of respondents in Qventus’ survey ranked the pressure to operationalize AI as seven or higher on a scale of one to 10.
Measuring return on investment also remains a challenge for systems, with four of five respondents reporting difficulty measuring the ROI of AI. Thirty-nine percent report not having a clear benchmark for performance or ROI.
Thirty-eight percent of respondents also say they do not expect to see ROI for 13 or more months, though 74% report needing it within a year or less to justify an AI investment.
Among the 60 CIOs, chief AI officers, CMIOs and other senior leaders that Qventus surveyed, 42% report actively deploying AI across multiple use cases, but only 4% have achieved scaled AI implementation with measurable outcomes. Nearly a quarter report that their piloting solutions operate in limited areas.
"If you make a wrong bet on a technology, you can blow your entire margins. Not-for-profit healthcare makes grocery stores look like they have excellent margins," said James Whitfill, M.D., Chief Transformation Officer at HonorHealth, in the report.
Another obstacle for systems is limited IT resources when managing multiple AI vendors, with 66% of respondents citing it as an issue. Twenty-five percent of respondents reported managing four to seven AI vendors. About half of survey respondents report spending 11% to 25% of IT bandwidth on vendor management, integrations and implementations alone, and 17% report spending as high as 26% to 50% of their organization's IT bandwidth.
“Working with too many AI vendors creates complexity in the system and produces additional, unnecessary costs,” the report said. “This problem is not new: a couple of years ago, businesses confronted it with the proliferation of AI assistants whose usage exploded across enterprises and later needed to be reined in.”
Despite the obstacles, 94% of respondents said delaying AI operationalization would put organizations at a “competitive disadvantage” against other systems, with 68% of respondents saying such delays would also increase clinician burnout and turnover.
“There is certainty that AI will dramatically reshape health systems in the years to come,” the report said. “But at what pace, at what price, and at what scale is still to be determined."