Baurzhan Ibrashev/istock via Get

News

R1 adds AI utilization management for denial prevention

The company's new Phare Utilization Management tool layers payer-specific intelligence onto a clinical AI engine to identify medical necessity denials in real time.

Jacqueline LaPointe
By
Published: 17 Sep 2026

R1 RCM is expanding its Phare Operating System platform with a utilization management capability that combines clinical and documentation data with payer-specific intelligence to identify cases that may be vulnerable to medical-necessity denials in real time.

The tool, dubbed Phare UM, continuously evaluates clinical evidence alongside payer-specific risk, documentation gaps and denial vulnerability, according to the announcement. It is designed to prioritize cases while the patient is still in the hospital, rather than discovering the problem retrospectively during denial management.

This moves UM closer to the front end of denial prevention.

UM traditionally sits at the intersection of clinical appropriateness and documentation, while payer behaviors are often managed downstream through denials and appeals. But R1's latest update to its platform aims to integrate these functions into a single workflow, using AI agents for clinical judgment and denial risk assessment.

R1 also said the tool can generate recommendations based on the company's proprietary data.

The approach could reduce the handoff problem, in which UM identifies an issue while another team handles denials. Instead, a single workflow supports denial prevention by enabling UM to not only assess whether patients meet payer criteria for payment but also what a particular payer might do with the case.

"By surfacing the right clinical and payer insights at the right time, Phare UM helps clinicians spend less time gathering information and more time making complex statusing decisions," Dr. Jennifer Weinberg, VP of physician advisory solutions at R1, said in the announcement.

Extending R1's pre-bill architecture

Phare UM is not a tool in isolation, but part of R1's pre-bill architecture through Phare OS.

R1 describes Phare OS as healthcare's first revenue operating system that connects clinical records and financial data into a single platform, while using AI agents to automate key functions and speed up payments. The platform stemmed from R1's acquisition of Phare Health in October 2025, which brought generative AI capabilities for clinical documentation improvement and complex inpatient coding.

R1's vision for the platform is to shift revenue cycle management from back-end remediation to front- and mid-cycle prevention. Phare UM brings that vision to life, according to CEO Joe Flanagan, by pinpointing denial risk sooner and closing documentation gaps before payers spot them.

The tool does this by tapping into R1's clinical AI engine and its proprietary foundational intelligence network within Phare OS, called Payer Atlas. This acts as a "smart connectivity fabric," R1 says, to track, map and decode complex, changing payer behaviors and rules. It also leverages the company's clinical AI engine within the platform.

Creating an automated, single workflow for UM and denials prevention also enables the review of all patient cases from discharge and surfaces high-need cases to UM teams, rather than relying on a queue based on admission order, the company explained.

Building an autonomous revenue cycle

This is yet another move toward a more autonomous revenue cycle.

Revenue cycle management has been a top use case for automation in healthcare, but the process as a whole has remained human-led. As AI evolves, it brings an opportunity for technology to take the lead, continuously detecting risk and recommending or initiating the next action. Humans then intervene when their expertise is needed.

R1's Phare UM may offer a glimpse of where the revenue cycle is headed as vendors increasingly jump on the autonomous bandwagon.

Waystar, for example, has made the autonomous revenue cycle an explicit aspect of its product strategy, describing an agentic network that can use clinical, financial and administrative data to determine and execute actions within workflows.

The goal is no longer simply to automate tasks within the revenue cycle, but to give technology enough context to identify problems, determine what should happen next and execute at least some of that work without a person managing every step.

Still, autonomous revenue cycle technology has much to prove. There remains a gap between predicting a denial and preventing one, and providers will need to determine if more intelligence can close the loop.

Jacqueline LaPointe is an Executive Editor at Xtelligent Healthcare Media, covering revenue cycle management, healthcare payers, health policy and health IT since 2016.

Dig Deeper on Claims reimbursement