Agentic AI Could End Healthcare's Payer-Provider Arms Race

Overview

The healthcare industry is currently locked in an accelerating administrative arms race between providers and payers. In 2026, artificial intelligence tools have already been deployed to draft prior-authorization requests for physicians and to scrutinize those same requests for health plans. When a claim is denied, additional AI systems are now available to construct appeals.

However, this automation has created a paradox: while the speed of dispute resolution has increased, the underlying friction remains. We have built intelligent tools for every stage of disagreement but failed to build systems that prevent the disagreement from occurring in the first place. The patient, ultimately, continues to wait.

Agentic AI—the next evolution of systems capable of interpreting conditions, consulting multiple sources, and coordinating actions—presents a critical inflection point. It offers the potential to shift the focus from automating arguments to resolving the misalignment that causes them.

The Convergence of Regulation and Technology

This moment is distinct because regulatory pressure, interoperable data standards, and advanced AI capabilities are arriving simultaneously. New requirements from the Centers for Medicare & Medicaid Services (CMS) are forcing a change in how prior authorizations are handled. By January 2027, affected payers must implement new application programming interfaces (APIs) to improve information exchange among patients, providers, and health plans.

These regulations also demand shorter decision timelines and more specific reasoning for denials. Major insurers have pledged to reduce authorization requirements, standardize electronic submissions, and move toward real-time approval decisions. Simultaneously, AI is moving beyond simple text generation. A new class of agentic systems can now coordinate complex sequences of actions toward a defined goal, rather than merely responding to isolated prompts.

This convergence creates a strategic choice for the industry. Stakeholders can continue to use AI to help providers submit more requests while helping payers scrutinize them more efficiently, thereby accelerating the administrative arms race. Alternatively, they can change the target, using these technologies to address the root causes of payer-provider friction.

Misalignment, Not Just Paperwork

Payer-provider friction is often characterized as an administrative burden, but this description is incomplete. The deeper issue is a fundamental misalignment of information and objectives. Clinical evidence and payer requirements often live in separate systems, creating a situation where policy criteria may only become visible to a provider after a request has already been submitted.

Each organization optimizes its own part of the transaction, even when doing so generates additional work across the system. Consequently, a denial is rarely just a failed claim; it is often a delayed signal that two organizations were operating from different information, rules, or assumptions.

Traditional Revenue Cycle Management (RCM) is built to respond to this signal after the fact. It focuses on correcting claims, assembling appeals, scheduling peer-to-peer reviews, and pursuing payment. This approach treats friction as a problem to be managed rather than a symptom to be prevented.

A more intelligent revenue cycle would treat friction as information and move that information upstream.

From Speed to Orchestration

At first glance, the solution appears to be speed. If a prior authorization traditionally takes fourteen days, the immediate impulse is to build a system that processes it in one. However, speed alone does not resolve the underlying problem. If a provider submits an incomplete request in seconds and a payer denies it in seconds, the industry has merely achieved real-time dysfunction.

Traditional automation asks a simple question: How can I complete this task faster? Agentic orchestration asks a more complex set of questions: Why did this task become necessary? What information is missing? Which action should happen next?

This represents a significant architectural shift. The goal is no longer to automate isolated steps, such as checking eligibility or generating an appeal. Instead, the objective is to coordinate the entire process so that the right information reaches the right system before a preventable failure occurs.

The Architecture of Agentic Care

The breakthrough is not a swarm of chatbots sending messages to one another. It is an orchestration layer built on top of healthcare’s emerging interoperability infrastructure. Consider a scenario where a physician orders a knee MRI in an electronic health record (EHR). The order is represented using FHIR, a standard for exchanging healthcare data through APIs.

Three implementation guides developed through the HL7 Da Vinci Project can support the authorization process:

Coverage Requirements Discovery (CRD) asks the payer whether authorization is required and which coverage rules apply.
Documentation Templates and Rules (DTR) identifies the required evidence and helps retrieve it from the patient’s record.
Prior Authorization Support (PAS) packages the request, sends it to the payer, and returns the response.

These standards create the rails, while agentic AI coordinates what travels across them. An orchestration agent could maintain the state of the request and determine which service to call next. A retrieval agent might search both structured data and unstructured clinical notes for the patient’s physical therapy history. A policy engine would then compare that evidence with the payer’s current criteria.

Crucially, the language model should help interpret messy clinical information, but it must not become the source of truth. If an agent claims that a patient completed six weeks of physical therapy, it should link that claim to exact encounters or notes within the medical record. This distinction ensures that efficiency does not come at the cost of accuracy.

By integrating these components, the industry can move from a model of automated disagreement to one of coordinated care, where the relationship between payers and providers is repaired rather than merely expedited.