Opportunity and risk of embedded EHR AI

 The next phase of the EHR is arriving quickly, and it is no longer just about documentation.

Major EHR vendors have accelerated AI rollouts in recent weeks. On Feb. 2, Oracle Health expanded its Clinical AI Agent to automate order creation for lab tests, imaging, prescriptions, referrals and follow-up appointments during patient visits. Two days later, Epic released AI Charting, a tool that listens during visits, drafts notes and suggests orders in real time as part of its ART AI suite.

CIOs say embedding AI directly into the EHR could finally give clinicians meaningful time back.

“AI that isn’t embedded in the EHR is like having a capable assistant who sits outside the room,” Muhammad Siddiqui, CIO of Reid Health in Richmond, Ind., told Becker’s.

When AI is embedded, he said, it becomes part of the work instead of a separate task. It can listen, suggest and act within the same workflow clinicians already trust. 

“That’s where the real opportunity lies. Less context switching. Fewer handoffs. More consistency across teams,” Mr. Siddiqui said. “Done well, it helps health systems reduce variation and support clinicians without asking them to work harder.”

Darrell Bodnar, CIO of North Country Healthcare in Whitefield, N.H., echoed that sentiment, saying that as AI becomes embedded directly into the EHR, the biggest opportunity for health systems is reducing cognitive and administrative burden — particularly through ambient documentation, workflow automation and intelligent summarization — to give clinicians time back without altering clinical judgment.

But CIOs caution that embedded does not automatically mean better.

“The risk is thinking embedded automatically means better,” Mr. Siddiqui said. “If the AI is noisy, inconsistent or hard to govern, it can quietly undermine trust. Clinicians are quick to disengage when tools feel unreliable or create rework.”

There is also risk in the speed of deployment. EHR vendors are rolling out generative AI-powered features faster than many health systems can absorb, said J.D. Whitlock, CIO of Dayton (Ohio) Children’s in creating the potential for “analysis paralysis” in AI governance or rushed decision-making.

For some leaders, the concern extends beyond governance to vendor control.

James Wellman, vice president and CIO of Nathan Littauer Hospital & Nursing Home in Gloversville, N.Y., said AI built into the EHR can create integration barriers for third-party solutions.

“In my opinion, [EHR vendors] are hesitant to allow some of these vendors into their market share,” he said. “That creates an initial challenge for us and vendors to overcome before we can even start.”

His rural health organization is developing a unified AI approach that spans call center agents, search and summarization across current and archived data, revenue cycle, and ambient AI tools. However, capabilities, he said, may not always align with broad, built-in solutions.

“We are a rural health organization, and the broad, generalized, built-in solutions may not fit our needs,” Mr. Wellman said. “Third-party vendors are working for us every day to keep our trust.”

For decades, the EHR functioned largely as a digital archive, essential but reactive. As AI becomes embedded into the EHR, that system of record is evolving into something more active, one that can listen, suggest and increasingly shape workflows in real time.

For CIOs, the promise is compelling. But as AI moves deeper into the core of clinical systems, they say the challenge will be capturing the benefits without compromising trust, governance and control.

https://www.beckershospitalreview.com/healthcare-information-technology/ehrs/cios-weigh-opportunity-and-risk-of-embedded-ehr-ai/