For decades, prior authorization has been one of the most quietly expensive problems in American healthcare. A physician orders a treatment, and before a patient can receive it, a health plan or pharmacy benefit manager (PBM) has to confirm it’s medically necessary and covered. On paper, it’s a compliance checkpoint. In practice, it’s often a fax machine, a phone queue, and a multi-day wait, a delay that can mean the difference between timely care and a worsening condition.
The American Medical Association has repeatedly flagged prior authorization as a leading driver of physician burnout and care delays, with surveys showing the average practice completes dozens of authorization requests per physician, per week, largely through manual or
semi-manual processes. Health plans face the mirror problem: reviewing that volume of requests consistently, compliantly, and fast enough to meet state and federal turnaround mandates, without over-hiring clinical reviewers to do it.
That gap, between the volume of decisions that need to be made and the manual capacity to make them, is exactly where AI and intelligent automation are starting to change the equation.
From Paper Rules to Real-Time Decisions
The first generation of utilization management (UM) software largely digitized paper: it moved forms online but still relied on staff to read a request, check it against a policy manual, and make a judgment call. The newer generation of platforms is different: they encode clinical criteria directly into rules engines that can evaluate a request against medical necessity guidelines the moment it’s submitted, flagging only the genuinely ambiguous cases for human review.
This is the model companies like Agadia have built their platforms around. Agadia, a healthcare technology company serving health plans and PBMs, positions its software explicitly as
“AI-powered automation for modern utilization management,” covering electronic prior authorization, medication therapy management, formulary benefit design, and grievance management for organizations processing millions of authorization decisions a year. The goal isn’t to remove clinical judgment from the process; it’s to reserve it for the cases that actually need it, while routine, clearly-compliant requests move through in minutes instead of days.
It’s worth noting that prior authorization is only one piece of the utilization management workload. Formulary benefit design, medication therapy management, and grievance handling all carry similar administrative weight, and all involve the same underlying pattern: a large volume of decisions that need to be evaluated against defined criteria, tracked for compliance, and documented in a way that holds up to audit. Platforms that apply automation across that full set of functions, rather than prior authorization alone, tend to see the efficiency gains compound, since the same rules-engine infrastructure that speeds up one workflow can usually be extended to the others with far less incremental effort than building each one from scratch.
Why This Matters Beyond Efficiency
It’s tempting to frame this purely as a cost or speed story, but the more important shift is in accuracy and consistency. A manual review process, run across dozens of reviewers, inevitably produces variation: the same request can get approved by one reviewer and questioned by another, simply because of how consistently policy criteria are applied. Automated decision support built on structured clinical rules reduces that variability, which matters both for compliance audits and for patient trust in the system.
There’s also a growing regulatory dimension. With CMS finalizing rules requiring faster electronic prior authorization turnaround times and greater transparency into denial reasons for Medicare Advantage and other federally regulated plans, health plans that are still running UM on manual or legacy systems are facing a hard deadline to modernize. AI-driven automation isn’t just a competitive advantage anymore. For many plans, it’s becoming a compliance requirement.
The cost side of this shouldn’t be understated either. Every manual review touches staff time, and staff time scales linearly with request volume in a way that automated review doesn’t. As pharmacy and medical benefit utilization continues to grow, plans that rely on adding headcount to keep pace will find that approach increasingly expensive relative to plans that have already shifted the routine share of that workload onto software. That gap in operating cost tends to widen over time rather than narrow, which is part of why modernization has moved from a discretionary IT project to a budget priority in a lot of organizations.
What Responsible Adoption Looks Like
The health plans getting this right aren’t treating AI as a black box that approves or denies care unsupervised. The more durable model keeps a human reviewer in the loop for anything outside clearly defined clinical criteria, uses AI to handle the high-volume, low-ambiguity decisions, and keeps a full audit trail so every automated decision can be explained and defended. That combination, automation for scale and human judgment for nuance, is what separates a UM program that survives a compliance audit from one that doesn’t.
The Road Ahead
Prior authorization isn’t going away. It’s a structural part of how U.S. healthcare manages cost and appropriateness of care. But the manual, fax-and-phone version of it is increasingly unsustainable, for physicians and health plans alike. As platforms built for automated,
rules-based, auditable decision-making continue to mature, the plans and PBMs that adopt them early are likely to see the benefit twice over: faster care for patients, and a utilization management operation that can actually keep pace with regulatory expectations.
Health plans and PBMs looking to modernize prior authorization, formulary management, and utilization review can learn more about Agadia’s AI-powered automation platform at agadia.com.
