The electronic health record has spent decades as healthcare’s most indispensable and most complained-about system — simultaneously the foundation of clinical documentation and the primary source of clinician administrative burden. At HIMSS 2026, the major EHR vendors arrived in Las Vegas with a unified message: that equation is changing. Across announcements from Epic, Oracle Health, MEDITECH, athenahealth, and Greenway Health, the EHR is being repositioned not as a system of record but as an autonomous operating platform, one that can draft clinical notes, manage prior authorization, appeal denied claims, and open its data to a new generation of AI agents.
HotSpot Take
At HIMSS26, EHR vendors stopped positioning AI as a feature and started presenting it as infrastructure — with outcomes data, native architectures, and interoperability moves that signal a fundamental shift in what an electronic health record is expected to do.
Epic: Outcomes Data and a Development Platform
EHR vendors at HIMSS26 presented outcomes data and new AI capabilities that reframe the electronic health record as an active participant in clinical and operational workflows. Photo courtesy of HIMSS26.
No EHR vendor commanded more HIMSS26 attention than Epic Systems, which arrived with something more persuasive than product announcements: a portfolio of documented outcomes from live deployments at named health systems.
The headlining announcement was Agent Factory, a no-code visual builder embedded within the Epic environment that allows health systems to design, configure, and deploy custom AI agents tailored to their own workflows, policies, and knowledge bases. The significance of Agent Factory is structural: Epic is not prescribing which agents health systems should use — it is giving them the tools to build their own, within the EHR environment, on their own timeline.
That announcement was framed by performance data from Epic’s existing AI suite, which includes Art (clinical documentation), Penny (revenue cycle), and Emmie (patient engagement). According to Epic, more than 85 percent of its customer base now actively uses one or more of these tools. The outcomes figures cited at the conference were specific and attributed to named organizations. At Riverside Health in Virginia, clinicians using Art’s inpatient insights reduced time spent on documentation and communication tasks by up to 32 percent. Discharge summaries are being completed 20 to 30 percent faster across multiple organizations using Art’s Draft Hospital Course Notes. At The Christ Hospital, Art’s analysis of incidental radiology findings has driven a 69 percent early lung cancer detection rate, compared to a national average of 46 percent, according to Epic.
On the revenue cycle side, Penny’s results were sharper still. At Summit Health, medication prior authorization submission time dropped 42 percent, with 92 percent of AI-generated responses accepted without edits. At Rush University Medical Center, Emmie delivered a 58 percent reduction in billing-related customer service messages.
Epic also unveiled Curiosity, a proprietary family of medical foundation models trained on anonymized real-world patient records, designed to predict clinical events across diagnoses, medications, procedures, and outcomes. The models are intended to extend Art’s capability from documentation toward active clinical decision support.
Oracle Health: Expanding into the Emergency Department
Oracle Health brought a focused announcement to HIMSS26: its Clinical AI Agent note generation capability is now available in U.S. inpatient and emergency department settings, extending a solution that had previously been limited to ambulatory care.
The expansion is clinically meaningful. Emergency department documentation presents distinct challenges — encounters are fast-paced, multi-event, and often involve multiple clinicians interacting with the same patient across a single visit. Oracle’s Clinical AI Agent addresses this by drawing on triage notes, initial and re-examination data, lab results, imaging, and prior notes to draft a single comprehensive encounter note that reflects the full arc of the patient visit. Clinicians review, edit, and sign the draft rather than building the note from scratch.
AtlantiCare, the New Jersey-based health system, expanded its deployment of the Clinical AI Agent to all emergency departments after initial success in ambulatory settings, reporting a 41 percent reduction in documentation time. According to Oracle, the agent has now saved clinicians more than 200,000 hours across all U.S. users.
“Today more than 300 organizations are using the agent, which has helped save doctors more than 200,000 hours across all users.” — Seema Verma, EVP and General Manager, Oracle Health and Life Sciences
Seema Verma, executive vice president and general manager of Oracle Health and Life Sciences, described the company’s broader direction at HIMSS26: “Today more than 300 organizations are using the agent, which has helped save doctors more than 200,000 hours across all users. This also enables clinicians to care for more patients each day, which can generate increased revenue for their clinics.”
Oracle also highlighted its next-generation EHR platform, currently available for ambulatory providers in the U.S. and expected to reach acute care settings later in 2026. The platform embeds AI directly into clinical workflows and is built on an open architecture that allows organizations to extend Oracle’s AI agents, build their own, or integrate third-party models within a secure, patient-centered environment.
MEDITECH: Native Ambient and Autonomous Denial Agents
MEDITECH used HIMSS26 to debut two meaningful additions to its Expanse platform that reflect the same native-AI direction as its larger competitors — without the scale of Epic’s announcement or Oracle’s deployment numbers, but with a clarity of purpose that resonates with the community hospital and mid-market segments MEDITECH serves.
The first was native ambient listening tools built directly into Expanse’s point-of-care applications. Rather than relying on a third-party ambient scribe integrated into the EHR workflow, MEDITECH’s approach embeds the listening and documentation capability within the platform itself — reducing integration complexity for health systems that have resisted adding yet another vendor relationship to their clinical environment.
The second was Claim Denial Agents: autonomous tools that research the root cause of a denial and draft evidence-based appeals using the patient’s chart. The agents operate within the EHR workflow, meaning the denial management process begins in the same environment where clinical care is documented — closing the loop between care delivery and reimbursement without requiring staff to navigate a separate revenue cycle application.
athenahealth: Interoperability as Competitive Positioning
athenahealth took a different angle at HIMSS26, centering its announcements on interoperability and data access rather than agentic automation — a positioning that reflects both the ambulatory market it serves and a bet that data infrastructure will determine which AI applications succeed in practice.
The centerpiece was athenaConnect, a new intelligent interoperability layer designed to serve as a single access point connecting athenahealth’s 170,000 providers — representing care delivery to approximately 20 percent of the U.S. population — to health systems, pharmacies, labs, diagnostic services, and external partners. According to the company, athenaConnect aims to move interoperability beyond regulatory compliance and into real-time clinical utility, making relevant data from external sources actionable within the athenaOne workflow at the point of care.
The more technically forward-looking announcement was a preview of athenahealth’s industry-first patient-facing Model Context Protocol (MCP) server, which enables patients to grant authorized AI agents secure, structured access to their athenaOne health data. The capability is designed to allow AI tools to generate personalized clinical insights using a patient’s longitudinal health record while maintaining patient control over data permissions. The MCP server represents a foundational move: as AI agents proliferate, the ability to securely access structured EHR data — rather than relying on unstructured summaries — will determine the depth of clinical utility those agents can deliver.
athenahealth also demonstrated a Patient Conversations feature enabling practices to initiate secure text chats with patients directly from the EHR, with AI-powered handling of patient-initiated calls and staff escalation capability.
Greenway Health: Building AI Into the Foundation
Greenway Health made the most structurally ambitious EHR announcement at HIMSS26 with the launch of Novare, a fully integrated clinical and revenue cycle platform built from the ground up around an AI engine rather than layering AI on top of a legacy database. The ambulatory-focused platform is designed to perform coding and prior authorization instantly during the visit, at the point of care, rather than as a post-encounter workflow.
According to Greenway, a 10-provider pilot practice reported saving approximately 14,000 hours per year through automated coding and prior authorization workflows. Novare positions itself as an alternative for ambulatory practices that have grown frustrated with the latency and complexity of AI capabilities layered onto legacy EHR architectures — an argument that resonates particularly in the independent and mid-market practice segments that Greenway serves.
PointClickCare and the Post-Acute Dimension
The EHR conversation at HIMSS26 extended beyond the acute and ambulatory settings to include the post-acute and long-term care market, where PointClickCare introduced Discharge Intel — an AI-powered solution designed to give health plans timely clinical intelligence within 24 hours of hospital discharge.
The announcement addresses a persistent care transition challenge: health plans often lack real-time visibility into the clinical status of members being discharged from acute care, creating delays in post-acute authorization, care coordination, and case management. By surfacing structured clinical intelligence at the point of discharge, PointClickCare positions Discharge Intel as a tool that simultaneously serves the provider’s need for faster transitions and the payer’s need for timely member data.
A Market Redefining Its Core Product
The EHR announcements at HIMSS26 collectively describe an industry in the process of redefining what its core product does. For the better part of two decades, the EHR has been primarily a documentation system — a place to record what happened during a clinical encounter. The announcements in Las Vegas suggest that the next chapter positions the EHR as an active participant in clinical and operational workflows: drafting notes, appealing denials, surfacing evidence, coordinating data access, and deploying AI agents on behalf of the health systems that host it.
That shift has implications beyond the EHR vendors themselves. As Epic’s Agent Factory, Oracle’s open architecture, and athenahealth’s MCP server each create new frameworks for how AI agents access and act on EHR data, the competitive dynamics for the broader health IT ecosystem are being restructured in real time.
Photo credit: HIMSS26
— This original article was created with AI support.
