For decades, prior authorization operated as one of American healthcare’s most consequential blind spots. Physicians submitted requests, insurers approved or denied them, and almost no one outside those transactions knew the aggregate numbers. How often did plans say no? How long did decisions take? How often did patients appeal and win?1,3
March 31, 2026 changed that. Under the CMS Interoperability and Prior Authorization Final Rule (CMS-0057-F), health insurers were required to post prior authorization metrics on their public websites for the first time. Approval rates, denial rates, appeal overturn rates, and processing times are now on the record.1
What’s emerging from that data is not a minor compliance footnote. It is a structural indictment of how prior authorization has functioned and a signal that CMS intends to redesign the system from the ground up rather than tinker at the margins.1,5
“CMS is not simply trying to reduce administrative burden. It is defining the operating model through standards, timelines, and data exchange.”2
Prior authorization was designed as a cost-control mechanism. In practice, it evolved into something physicians describe as the single greatest barrier between their clinical judgment and patient care.3
The American Medical Association has documented physicians managing an average of 39 to 40 prior authorization requests per week, consuming roughly 13 staff hours that would otherwise go toward patient care. The downstream consequences are significant: delayed diagnoses, abandoned treatment plans, and physician burnout are well-documented outcomes of a system that subjects time-sensitive clinical decisions to administrative review processes that can stretch for days.3,9 For payers, the overhead is equally costly, and the reputational exposure that was once invisible is now public.3,10
The rule is built on four requirements that, taken together, reframe prior authorization as regulated infrastructure rather than a payer-managed administrative process.1
Decision timelines tightened on January 1, 2026: urgent requests must be resolved within 72 hours and standard requests within seven calendar days, down from 14. Every denial must now include a specific clinical or coverage-based rationale. Critically, this applies to AI-assisted denials as well, a first-of-its-kind provision that places the burden of explainability directly on payers deploying machine learning in utilization management.1 Payers must also publicly disclose approval rates, denial rates, appeal outcomes, and processing times on an annual basis. The first reports, covering calendar year 2025, were due March 31, 2026.1
The final pillar, FHIR-based API implementation for electronic prior authorization exchange, carries a compliance deadline of January 1, 2027.1 That date is closer than most organizations’ technology roadmaps reflect.10
The rule continues to expand.4 In April 2026, CMS released a proposed rule that extends these reforms into drug benefits covered under both medical and pharmacy plans, with a public comment deadline of June 15, 2026. If finalized, the scope of CMS-0057-F will reach further than many organizations have planned for.4
The numbers now on public record are worth sitting with.5 KFF reported that Medicare Advantage insurers made 52.8 million prior authorization determinations in 2024, denying 4.1 million of them. That is a denial rate of 7.7%. Of those denials, only 11.5% were appealed, but when appeals were filed, 80.7% were overturned.5
If 80.7% of appealed denials are overturned, the question that follows is unavoidable: how many of the 88.5% that were never appealed should have been?5
The appeal paradox is the most consequential finding from the initial transparency filings. It points to a systemic over-denial problem that persisted largely because there was no mechanism for public accountability. That accountability now exists, and it is retroactive in its implications.5
Industry-wide, first-pass prior authorization approval rates sit between 70 and 80 percent. High-performing organizations are achieving 90 percent or higher through investment in documentation quality, submission process discipline, and technology. Under CMS-0057-F, that performance gap is no longer an internal metric. It is a public comparison.6
Artificial intelligence is reshaping prior authorization on two tracks that are moving in opposite directions.9
Commercial payers are deploying machine learning models to triage PA requests in near real-time, with the promise of faster decisions and lower administrative overhead.9 The clinical community’s response has been skepticism.7 An AMA survey from February 2025 found that three in five physicians believe AI is accelerating automated denials rather than reducing the burden.7 The concern is not that AI is being used. It is that AI may be optimizing for the same flawed outcome, just faster.7,9
CMS has stepped directly into this tension.1 Starting in 2026, payers must disclose the reasoning behind every AI-assisted denial, a mandate designed to prevent algorithmic decision-making from functioning as a black box within a newly transparent system. This is not a peripheral requirement.1 For any health plan or utilization management vendor deploying predictive models in their review process, explainability is now a regulatory obligation.1
The organizations that will lead this space are not the ones that simply automate existing workflows. They are the ones that can demonstrate their AI improves decision accuracy, reduces over-denial, and produces reasoning a clinician can evaluate and a regulator can review.1
For health plans, the immediate priority is understanding where your prior authorization performance stands relative to the data now publicly available. Denial rates, appeal overturn rates, and turnaround times are no longer internal metrics. They are benchmarks against which every payer in your market is being compared. Organizations that have not established internal baselines will find themselves reacting to external scrutiny rather than getting ahead of it.1,5
For utilization management technology vendors, FHIR PAS API readiness is the most pressing operational question. January 2027 is a hard deadline, and platforms that achieve compliance early will have a concrete differentiator in the next procurement cycle. Interoperability is not a feature. It is the cost of entry.1
For provider organizations, the path forward is building toward 90 percent or higher first-pass approval rates through documentation discipline and electronic submission readiness before the mandate takes full effect. The practices that do this early will see both operational and financial returns.6
Across all stakeholders, the CMS comment period for the proposed rule expanding CMS-0057-F to drug benefits closes June 15, 2026. This is a direct opportunity to shape what comes next by documenting the operational and clinical realities that regulators need to hear.4
CMS has handed the healthcare industry a forcing function.1,8 The voluntary pledge signed by roughly 60 health insurers in June 2025 to streamline prior authorization represents movement, but the AMA’s 2025 survey found only one in three physicians believe it will produce meaningful change.8 The organizations that will define what prior authorization looks like in five years are not the ones who meet the minimum requirements. They are the ones that recognize transparency, interoperability, and accountable AI as competitive advantages, not obligations.
The black box is open. What is inside it, and what your organization chooses to do with that information, will matter more than most healthcare leaders currently realize.
1. CMS.gov, CMS Interoperability and Prior Authorization Final Rule (CMS-0057-F): cms.gov/CMS-0057-F
2. AJMC, “Prior Authorization in 2026: CMS Is Rebuilding the Operating Model” (May 2026): ajmc.com/prior-auth-2026
3. American Medical Association, Prior Authorization Physician Burden Survey: ama-assn.org/prior-authorization
4. CMS.gov, Prior Authorization Proposed Rule, April 2026 (Comment period closes June 15, 2026): cms.gov/newsroom/fact-sheets/2026-cms-interoperability-standards-prior-authorization-drugs-proposed-rule
5. KFF, Medicare Advantage Prior Authorization Determinations, 2024
6. Insight Health, Prior Authorization Approval Rate Analysis, April 2026: insighthealth.ai/prior-auth-approval-rate
7. American Medical Association, Prior Authorization and AI Survey, February 2025
8. American Medical Association, Prior Authorization Reform Pledge Survey, 2025: ama-assn.org/prior-authorization-reform-pledge
9. Medical Economics, “Prior Authorization: How It Evolved, Why It Burdens Physicians, and the Promise of AI” (June 2026): medicaleconomics.com/prior-auth-ai
10. MGMA, The Prior Authorization Landscape in 2025: mgma.com/prior-auth-landscape-2025