Most health systems struggle to scale AI beyond pilots, with EHR dependencies cited as a top barrier

Just 4% of large health systems have scaled AI with measurable results, per Qventus research. EHR vendor dependencies and point-solution sprawl are the main barriers to broader deployment.

Categorized in: AI News Healthcare
Published on: Apr 12, 2026
Most health systems struggle to scale AI beyond pilots, with EHR dependencies cited as a top barrier

Healthcare Systems Struggle to Move AI Beyond Pilots, EHR Dependencies to Blame

Only 4% of large health systems have successfully scaled artificial intelligence implementations with measurable outcomes, according to research from Qventus released Thursday. The gap between AI experiments and operational deployments stems largely from electronic health record vendor dependencies and the complexity of managing multiple third-party software integrations.

Qventus surveyed and interviewed more than 60 senior healthcare IT leaders, including chief information officers and chief AI officers at medium and large national health systems. The findings reveal a field stuck in early stages despite widespread enthusiasm.

The Scaling Problem

Forty-two percent of respondents said their organizations were actively deploying AI across multiple use cases. But 45% cited "difficulty scaling pilots" as a major obstacle.

EHR vendor dependency emerged as the top execution barrier, cited by 74% of respondents. Health systems wait for AI feature rollouts from their EHR vendors while simultaneously managing implementations from 10, 15, or 20 separate point-solution vendors - a fragmentation that makes safe, scalable deployment nearly impossible.

A quarter of respondents reported lacking a clear process for benchmarking AI performance, indicating that many organizations cannot even measure whether their AI investments are working.

What Success Looks Like Now

Healthcare leaders are shifting how they measure AI success. Sixty-two percent now prioritize revenue generation, and 59% focus on hard dollar cost savings - a departure from earlier pilot metrics.

Half of respondents see potential in agentic and autonomous care platforms that handle scheduling, patient flow, and care gap identification with minimal human oversight. These systems could deliver higher returns than traditional decision-support tools.

Dr. Deepti Pandita, chief medical information officer and chief AI officer at UCI Health, said governance structures are evolving to accommodate this shift. "Today's governance is still human-in-the-loop, but tomorrow's may not include human-in-the-loop with advancements in autonomous AI," Pandita said.

The Cost of Waiting

Ninety-four percent of surveyed healthcare leaders said delaying AI deployment creates competitive disadvantage. Sixty-eight percent said it will worsen clinician burnout.

The pressure extends beyond technology. Federal spending cuts, workforce shortages, and demographic shifts - the population aged 65 and older is projected to increase 42% by 2050 - will strain healthcare delivery. "2040 may look worse than 1940 in terms of relative deprivation of access," said Dr. Joseph Sanford, chief clinical informatics officer at the University of Arkansas for Medical Sciences.

The Trust Problem

A separate study from EBSCO surveyed 1,000 clinicians and 1,000 consumers about AI-powered clinical decision support tools. Most clinicians said they use such tools multiple times per patient encounter and believe they improve outcomes.

Consumers tell a different story. While 89% of clinicians said AI-driven decision support leads to better outcomes, 64% of consumers would prefer to see a healthcare professional who does not use AI at all.

How Sutter Health Scaled Past Pilots

Sutter Health moved from small-scale pilots to enterprisewide AI implementations by maintaining three guardrails: validating efficacy using real-world data from their own patient population, using common AI infrastructure, and avoiding point-solution fragmentation.

"You can't integrate with 10, 15, 20 different point solution companies, separate vendors, and integrate them into your health system in a scalable and safe way," said Dr. Jason Wiesner, chair of Sutter's imaging service line.

The stakes are high. "If you make a wrong bet on a technology, you can blow your entire margins," said Dr. James Whitfill, chief transformation officer at HonorHealth. "Healthcare makes grocery stores look like they have excellent margins."

Learn more about AI for Healthcare and AI Agents & Automation implementation strategies.


Get Daily AI News

Your membership also unlocks:

700+ AI Courses
700+ Certifications
Personalized AI Learning Plan
6500+ AI Tools (no Ads)
Daily AI News by job industry (no Ads)