What payer use of agentic AI means for medical practices

As payers adopt AI, medical practices need to keep pace

Most of the talk about artificial intelligence in healthcare has revolved around its impact on provider productivity: the fact, for instance, that more than 80 percent of physicians say they’re using AI tools for documentation and administrative purposes.¹ Often missing from the conversation, on the other hand, is news about AI deployment among payers. Assuming that health plans are incorporating AI into their own workflows, how is the technology actually helping them? And for medical practices, what are the implications? Could payer AI change the reimbursement equation?

Insights into both questions were offered recently when the Deloitte Center for Health Solutions shared the results of its survey of 100 healthcare technology executives.² Deloitte asked the IT leaders, including 50 from insurance companies, to shed light on their use of agentic AI, a version of the technology that can take action and coordinate tasks with limited human intervention.³

The researchers found that 70 percent of surveyed health plans have deployed or have plans to deploy agentic AI for utilization management, prior authorization, and claims management purposes. The technology, noted one insurance executive in a separate online focus group, “monitors expiring licenses, verifies credentials against authoritative sources, proactively updates payer databases, and escalates exceptions for human review only when needed” (as reported by Deloitte).

The takeaway for medical practices is clear: The organizations monitoring claims and authorizing requests are becoming faster, more automated, and AI-driven. They’re rapidly operationalizing agentic AI across their administrative workflows, and that shift will increasingly shape how healthcare providers are paid. For smaller and independent practices, the question now is how to keep up. As the humans handling your claims work increasingly find that they’re grappling with new policies and payer-side machines operating at much faster speeds, what’s the solution for many practices that may not have unlimited technological resources?

Impact of payer agentic AI on practice revenue cycles

It’s easy to see why payers might be interested in applications for agentic AI. By automating administrative processes—reviewing PA requests, comparing submitted documentation against coverage policies, routing claims and identifying missing information—health plans can make decisions faster and more consistently than they might through traditional manual workflows alone.

For practices, though, that increased efficiency could create an operational mismatch in the realm of revenue cycle management. Lacking support from their own AI agents, they may find there’s less time to gather required documentation before submission, respond to requests for additional information, or resolve claims issues before reimbursements are delayed.

Among other things, practices may struggle with: 

• Prior authorization requests that receive faster responses, including requests for additional information
  • Potential practice impact: increased documentation requirements and resubmission requests, delaying care and increasing staff workload
• Claims that are adjudicated more quickly and more consistently
  • Potential practice impact: denials or requests for corrections arrive sooner, creating pressure to respond promptly to avoid reimbursement delays
• Appeals that are more documentation-intensive
  • Potential practice impact: additional staff time may be needed to locate records, compile evidence, and manage appeals

Overall, practices may have to spend more time reacting to payer requests than preventing claims issues from happening in the first place. (In 2024, 60 percent of medical group leaders reported that claim denials increased year over year;⁴ since then, denial rates have continued to rise.⁵) Similarly, evolving payer requirements could become increasingly hard to track, and finally, practices could face significant staffing challenges as billing and other administrative pressures increase.

The agentic AI RCM opportunity

The good news is, it’s not all doom and gloom on the AI-enabled payers' front. The same Deloitte survey found that 80 percent of the health systems were also prioritizing agentic AI for their own purposes—and, importantly, that revenue cycle management was among their top use cases for the technology.

At one major health system, the Deloitte report noted, IT leaders had deployed AI agents to support RCM processes like eligibility and benefit verification and claims-related clinical information exchange.⁶ Health system participants also highlighted how agentic AI could drive back-office efficiency with “fewer manual handoffs between systems, faster process cycle times, and more robust and resilient operations,” and ultimately allow for “autonomous follow-through, end-to-end orchestration across systems, and the redeployment of staff to higher-skill work.”

With AI technologies that automate repetitive administrative work—and that can be easily scaled to absorb seasonal volume spikes or adapt to changes to health plan rules—it seems that health systems are taking a page directly from the payer playbook.

Healthcare organizations are using agentic AI to:

• Assemble documentation
• Monitor claim status
• Surface missing information
• Identify denial trends
• Prioritize staff work queues

Furthermore, an increasing number of organizations are leveraging “agent-to-agent” communication for exchanging structured data, clinical documentation, and authorization requests. Faster, more efficient, and less prone to mistakes than typical human-mediated exchanges, the A2A process entails direct interaction between payers’ and providers’ respective AI models. For small, independent practices, it’s an operating model that could pay dividends by drastically reducing administrative friction. It could also free up busy staff to focus on more valuable, patient-centered work.

How athenaOne® can navigate the agentic ecosystem

AI-native athenaOne offers practices with an integrated RCM platform custom-built for ambulatory care. By embedding AI agents into practices’ RCM workflows, the system automates much of the revenue cycle work to help providers keep pace with evolving payer processes:

• Scale and network intelligence: athenahealth processes 300M+ claims⁷ and sends 180M+ consolidated clinical documents (CCDs) to payers annually⁸, giving it visibility into payer behavior, denial trends, and adjudication patterns across the ambulatory market
• athenahealth's RCM roadmap includes live and in-development agentic AI capabilities for remittance downloads, portal logins, claim status checks, portal appeals, and prior authorization status checks
• Other AI-powered athenaOne functions generally available: Automated Coding Denials Advice, Network Denial Anomaly Detection, Payer Policy Surveillance, Automated Insurance Selection, Authorization Determination and Success Prediction (add-on solution), Automated Medical Coding (add-on solution)

Preparing practices for the agentic-AI future

A few years ago, the 2023 KLAS RCM Summit found that claims, claims statusing, prior authorization, and coding were among the top areas where practices were seeking automation ROI—exactly the same areas that payers are now automating with help from agentic AI.⁹

For practices that have embraced automation and the value it provides, the logical next step is to find an RCM partner that understands the importance of this trend. With an AI-enabled RCM platform aiding with back-office functions, they should find it’s easier to keep up with health plan requirements with the potential to be paid faster and reduce burden on staff.

Practices, in other words, don’t have to become experts on agentic AI, or on the health plans that are increasingly using it. What they do need is technology that provides network intelligence, scalable automation, workflow orchestration, and adaptability to evolving payer requirements. Their RCM solution should be:

• already deploying AI agents on their behalf
• learning from and adapting to payer changes in real time
• building the infrastructure for A2A communication before it becomes reimbursement table stakes

Payer utilization of agentic AI is no longer on the horizon, it’s here now and increasing by the day. As major health systems adapt to this new world with agentic AI solutions of their own, ambulatory practices need RCM tools and services designed to keep them up to speed as well. See how athenahealth is building AI-powered revenue cycle tools designed for the way ambulatory practices actually work.