Epic is turning its EHR into an AI command center, and CIO plans just got tighter

Epic used UGM 2026 to send a blunt message to health system IT teams: the EHR is no longer just where clinical data lives, it’s where the next wave of automation will be built, deployed, governed and measured.

At the company’s state fair-themed Users Group Meeting in Verona, Wisconsin, Epic leaders leaned hard on speed and follow-through. Epic founder and CEO Judy Faulkner said the company tracked 167 projects announced at UGM 2025, delivered 84 on time, has seven completed and headed for special updates, and has 76 in progress and on track, according to Healthcare IT News. Fierce Healthcare reported the same project counts and highlighted Faulkner’s conclusion that what’s announced at UGM is “very likely to happen.”

For CIOs and operations leaders, that delivery cadence is the operational story. When a core platform ships AI features at conference tempo, the bottleneck moves to validation, change control, clinical governance, and training capacity, not whether the vendor can deliver.

Epic’s AI stack is consolidating around Ergo, Art, Emmie, and Penny

Epic framed its AI work around Ergo, a platform Faulkner first announced in 2025 and described this year as “here,” according to Healthcare IT News. Epic positioned Ergo as an approach that starts with human judgment and uses generative AI to find patterns and iterate quickly, built on Epic’s data assets and workflows.

Healthcare IT News tied Ergo’s foundation directly to Cosmos, Epic’s analytics and research platform, which Epic said contains 320 million deidentified patient records drawn from 23 billion patient encounters. Fierce Healthcare, citing an Epic spokesperson, separately reported Epic said more than 325 million patients have medical records in Epic, and that the software is installed in more than 3,700 hospitals.

Epic also emphasized that its copilots are meant to work as a system. Fierce Healthcare reported Epic is weaving generative AI into clinical, operational, and administrative workflows, with Art for clinicians, Emmie as a patient-facing assistant in MyChart, and Penny aimed at revenue cycle tasks, all presented as components that can function together as enterprise AI for healthcare.

Adoption was repeatedly cited as an important proof point. Healthcare IT News reported that more than 60 organizations are using Chart with Art, and that clinician and nurse summaries generated by Art are deployed across more than 300 health systems. Fierce Healthcare likewise reported use at more than 300 organizations and pointed to another Epic figure: in an Epic community survey, 68% of respondents said that visit summaries containing Care Everywhere notes surfaced new patient details they otherwise would have missed.

Agent Factory pushes “build it inside Epic” into 2027 planning

The biggest forward-looking takeaway for IT leaders was Epic’s Agent Factory. Forbes reported that Epic is framing Agent Factory as a no-code way for health systems to build their own AI agents, with broad availability targeted for 2027. As Forbes put it, the message is platform gravity: keep innovation within Epic instead of bypassing the EHR with outside tools and integrations.

That changes budgeting and operating models. If agent development happens “in the platform,” the line items shift from third-party subscription negotiations to internal capacity planning: who builds agents, who tests them, how token or usage costs are allocated, and who signs off when a locally-built agent changes clinical workflow.

Forbes also framed the tradeoff for CIOs as timing and ROI uncertainty. Even if no-code reduces technical lift, health systems still have to account for staff hours, governance overhead, and the economics of model usage, especially as agent concepts move from pilots to production.

Revenue cycle and access workflows are being pulled into the same release train

Epic’s AI story at UGM wasn’t confined to documentation support. It also went after the messy edges of access and revenue cycle, where “another portal” is often the actual day-to-day pain.

Healthcare IT News reported Epic is rolling out a prior authorization API called Coverage Requirements Discovery (CRD) designed to let clinicians check payer requirements inside workflow, rather than leaving the EHR for external portals or manual lookups. The outlet cited Ochsner Health as an early adopter and reported additional integrations at Froedtert ThedaCare Health, Denver Health, Summit Health, and payer partners including UnitedHealthcare, Network Health and Aetna. Healthcare IT News also reported 16 payers were testing CRD.

For revenue-cycle leaders, CRD is a measurable change, not a vision statement. If payer rules can be resolved at order entry, the operational benchmark becomes fewer rework loops: fewer staff touches to discover requirements, fewer incomplete orders that boomerang back, and fewer delays that show up as avoidable denials later.

Epic also pointed to operational impact data from existing AI assistants. Fierce Healthcare reported Ochsner said Ask Emmie saved more than 1,000 call center staff hours over 10 months, and that Community Health Network reported 73% fewer billing-related InBasket messages after going live in July. Healthcare IT News reported Tampa General, Ochsner Health, Penn Medicine and Sutter Health have integrated Ask Art, and that Sutter was first to go live with Ask Emmie in MyChart.

The practical implication is that AI adoption is moving from clinician productivity stories to contact center throughput and message-volume management. That’s where AI’s cost case can be audited inside existing service-level metrics, and where deployment decisions tend to move faster.

Security and “patching discipline” are now part of the AI conversation

Epic also used its main stage to talk security posture as AI expands. Healthcare IT News reported Epic Chief Security Officer Stirling Martin emphasized planning for patching and said Epic is participating in Project Glasswing, an AI cybersecurity initiative launched by Anthropic that includes about 150 organizations worldwide.

For operators, this matters because AI features in the EHR aren’t separate systems. They arrive as upgrades, configuration choices, and workflow changes. Security and risk teams will increasingly be asked to approve not only model behavior and data access, but also the speed at which AI-dependent features roll into production.

Where this lands in 2026, 2027 operating plans

• Update the governance map before the next upgrade: identify who owns clinical validation, model behavior review, and change control for Epic AI features arriving as standard releases (per Epic’s UGM delivery cadence reported by Healthcare IT News and Fierce Healthcare).
• Treat Agent Factory like a platform program, not a feature: start scoping internal “agent product owner” capacity, test environments, and cost allocation for model usage ahead of the 2027 wider release window reported by Forbes.
• For revenue-cycle and access leaders, put CRD in the payer integration roadmap: confirm which payers are live vs. in testing, what transactions are supported, and what success metric will be tracked (portal avoidance, turnaround time, or denial reduction) based on the CRD rollout described by Healthcare IT News.
• Tie patient-facing AI to contact center KPIs: Ask Emmie’s reported hour savings and message reduction (per Fierce Healthcare) suggest a clean measurement frame for MyChart assistant rollouts, especially for systems experiencing InBasket growth.