What is the CMS April 2026 proposed rule on drug prior authorization?

In April 2026, CMS proposed extending electronic prior authorization (ePA) requirements beyond non-drug items and services to include prescription drugs under both medical and pharmacy benefits. This expansion aligns ePA standards across EHR vendors, clinicians, and health plans, with CMS projecting $15 billion in savings over 10 years from reduced administrative burden and faster drug access for patients.

How much have carriers already reduced prior authorization requirements in 2026?

Leading health insurance carriers have already eliminated 11% of prior authorizations for services since their June 2025 pledge to reduce administrative burden. UnitedHealthcare went further, announcing elimination of prior authorization requirements for 30% of healthcare services in May 2026. These voluntary reductions signal the industry direction — but practices still need AI automation to handle the remaining 70-89% of PA requirements efficiently.

Why is AI infrastructure better than point solutions for prior authorization compliance?

AI infrastructure — as opposed to point solutions — handles both services-based and drug-based prior authorization through a unified platform. When CMS extends ePA requirements to prescription drugs, practices using point solutions need a separate tool for drug PA. AI infrastructure platforms like BAM AI treat services PA and drug PA as workflows within the same system, sharing patient context, payer intelligence, and documentation across both. UCI CMIO Dr. Deepti Pandita noted in May 2026 that 'if organizations start looking at AI as infrastructure and not as point solutions, ROI will follow.'

How does AI automation help practices prepare for the 2027 ePA mandate expansion to drugs?

AI automation prepares practices for the expanded 2027 ePA mandate by building the operational foundation now: automated formulary checking against payer-specific drug lists, real-time step therapy and quantity limit validation, clinical documentation assembly for medical necessity, multi-channel submission via FHIR APIs and payer portals, and status tracking across both services and drug PAs. Practices that automate services PA today can extend the same AI infrastructure to drug PA with minimal additional configuration when the rule takes effect.

CMS Expands Prior Auth Reform to Prescription Drugs: AI Automation for 2026-2027 Compliance

Your practice just figured out how to handle prior authorization for services. Now CMS wants to do the same thing for every prescription drug under both medical and pharmacy benefits — and the proposed rule dropped in April 2026.

The expansion isn't a surprise to anyone paying attention. CMS has been methodically building toward a unified electronic prior authorization (ePA) framework since the CMS-0057-F final rule. But for practices that are still running manual PA workflows — or worse, using one tool for services PA and a completely different process for drug PA — this proposed rule doubles the compliance surface overnight.

$15B
in projected savings over 10 years from extending ePA to prescription drugs (CMS, April 2026)

CMS Proposes Extending Electronic Prior Authorization to Drug Coverage

The April 2026 proposed rule extends electronic prior authorization requirements beyond non-drug items and services to include prescription drugs under both medical and pharmacy benefits. According to reporting by AJMC (May 27, 2026), this expansion aligns ePA standards across EHR vendors, clinicians, and health plans — creating a single technical framework for all prior authorization activity.

CMS projects $15 billion in savings over 10 years from this alignment, according to analysis published by PCG Health Policy on May 27, 2026. The savings come from three sources:

Reduced phone and fax burden: Drug PA currently involves more manual phone calls and fax submissions than services PA because electronic channels are less standardized for pharmacy benefits.
Faster time-to-therapy: Patients waiting for drug prior authorization approvals face treatment delays that compound clinical outcomes and downstream costs.
EHR vendor alignment: A unified ePA standard means EHR vendors build one integration instead of maintaining separate workflows for services and drug PA — reducing software costs passed through to practices.

For healthcare administrators, the immediate question is operational: How do you handle twice the ePA compliance surface without doubling your PA staff?

$15 Billion in Savings: What the New ePA Drug Rules Mean for Practices

The $15 billion figure is a system-wide projection. What does it mean at the practice level?

Consider the math. A mid-size multi-specialty practice processes 200-400 prior authorizations per week. Currently, roughly 60-70% of those are for services (procedures, imaging, specialist referrals) and 30-40% are for prescription drugs (biologics, specialty medications, controlled substances, step therapy requirements).

The services PA side has been moving toward electronic submission for years, accelerated by the CMS 2027 FHIR-based ePA mandate. But the drug PA side? It's still largely manual for many practices. Pharmacy benefit managers (PBMs) have their own portals, their own forms, their own processes. A practice might use CoverMyMeds for some drugs, call the PBM directly for others, and fax documentation for specialty pharmacy authorizations.

When CMS standardizes drug ePA under the same framework as services ePA, practices that already automated their services PA workflow gain a massive advantage: the same infrastructure handles both. Practices still running manual workflows face a different reality — they need to build or buy automation for an entirely new category of PA volume.

11%
of prior authorizations already eliminated by leading carriers since June 2025 pledge (CMS blog / PCG Health Policy)

Carriers Are Already Moving: UHC Cuts 30% of PA Requirements

The regulatory push is meeting voluntary industry movement. Leading health insurance carriers have eliminated 11% of prior authorizations for services since their June 2025 pledge to reduce administrative burden, according to CMS data cited by PCG Health Policy.

UnitedHealthcare went further. In May 2026, UHC announced elimination of prior authorization requirements for 30% of healthcare services. That's not a pilot program or a promise — it's a live operational change affecting millions of covered lives.

What this means for practices is nuanced. Fewer PA requirements sounds like pure upside, but the operational reality is more complex:

The remaining PAs are harder: Carriers eliminate easy, low-risk PAs first. What's left are the complex, high-dollar authorizations — specialty drugs, advanced imaging, surgical procedures — that require more documentation, more clinical justification, and more follow-up.
Payer rules change faster: When carriers adjust which services require PA, practices need to update their workflows immediately. A procedure that required PA last month might not this month — and submitting an unnecessary PA wastes staff time, while missing a newly required PA triggers a denial.
Drug PA is the next frontier: Carriers reduced services PA first because that's where CMS pressure was focused. As CMS extends ePA requirements to drugs, expect similar pressure on drug PA — but until those requirements are eliminated, practices still need to process every one.

The practices that win this transition are the ones whose systems automatically track which payers require PA for which services and drugs — and adjust in real time as rules change. That's not a manual process. That's AI infrastructure.

Why AI Infrastructure Beats Point Solutions for PA Compliance

Dr. Deepti Pandita, CMIO at UCI, told Healthcare Finance News in May 2026: "If organizations start looking at AI as infrastructure and not as point solutions, ROI will follow." The insight is particularly relevant for prior authorization, where the CMS drug PA expansion exposes a fundamental weakness in point-solution approaches.

Here's the problem with point solutions:

Tool A handles services PA. It knows payer rules for procedures, imaging, and specialist referrals. It submits via EDI or payer portals. It tracks status.
Tool B handles drug PA. It connects to PBM portals, CoverMyMeds, specialty pharmacy systems. It manages formulary exceptions and step therapy requirements.
Tool A and Tool B don't talk to each other. They don't share patient context, payer intelligence, or documentation. When a patient needs both a procedure authorization AND a drug authorization for the same episode of care — which happens constantly in oncology, rheumatology, neurology, and gastroenterology — your staff manually coordinates between two systems.

AI infrastructure solves this by treating all prior authorization — services and drugs — as workflows within a unified platform. The same patient record, the same payer intelligence engine, the same documentation assembly system, the same status dashboard. When CMS extends ePA requirements to drugs, an infrastructure-based AI system adds drug PA as a new workflow type — not a new product.

The Pandita insight is backed by operational data: according to Healthcare Finance News / Qventus research from May 2026, 45% of CIOs are now the primary AI decision-makers at their organizations. CIOs think in infrastructure, not point solutions. They're buying platforms that scale across use cases, not individual tools that solve one problem each.

"If organizations start looking at AI as infrastructure and not as point solutions, ROI will follow." — Dr. Deepti Pandita, CMIO, UCI (Healthcare Finance News, May 2026)

Growing Clinician Confidence in AI for Prior Authorization

An AJMC/Cohere Health survey published May 30, 2026 reveals growing clinician confidence in AI-powered prior authorization. This is significant because clinician buy-in has historically been the adoption bottleneck — not the technology itself.

The confidence shift makes sense when you consider what AI actually does for the physician in the PA workflow:

Documentation assembly: Instead of the physician spending 15-20 minutes per PA pulling clinical notes, lab results, and imaging reports into a payer-specific format, AI agents assemble the complete documentation package automatically from the EHR.
Medical necessity framing: AI agents map the clinical documentation to payer-specific medical necessity criteria — translating the physician's clinical reasoning into the language the payer's review system expects.
Status transparency: Instead of the physician's office calling the payer to check PA status, AI agents track every pending authorization in real time and surface only the ones that need physician attention.

When the CMS drug PA expansion takes effect, clinician confidence in AI-powered PA becomes even more critical. Drug prior authorization involves clinical decision-making that services PA often doesn't — step therapy justification, formulary exception rationale, drug interaction documentation. Clinicians who already trust AI for services PA are far more likely to extend that trust to drug PA within the same system.

The Autonomize AI/AHIP webinar scheduled for June 24, 2026 further signals industry convergence on AI-powered PA modernization, with a focus on responsible AI governance — addressing the exact clinician concerns that slow adoption.

How BAM AI Automates Prior Authorization Across Services and Drugs

BAM AI built its prior authorization automation as infrastructure from day one — not as a services-only point solution that needs a drug PA add-on bolted to the side.

Here's what that means operationally:

Unified authorization engine: Services PA and drug PA share the same patient context, payer intelligence, and documentation assembly pipeline. When a rheumatology patient needs both a procedure authorization for a joint injection and a drug authorization for a biologic, both flow through the same system with shared clinical documentation.
Automated formulary intelligence: AI agents check payer-specific formularies, step therapy requirements, and quantity limits before submission — flagging potential issues before they become denials.
Real-time payer rule tracking: When UHC eliminates PA for 30% of services, our system updates automatically. When a PBM adds a new step therapy requirement for a specialty drug, the system catches it at the point of prescribing.
Multi-channel submission: Whether the payer accepts drug PA via FHIR API, PBM portal, CoverMyMeds, or fax, BAM AI agents route each authorization through the optimal channel based on payer and drug combination.
FHIR-ready architecture: As CMS mandates FHIR-based ePA APIs for drug coverage, BAM AI's agents are already built on FHIR standards — meaning zero rearchitecting when the mandate takes effect.

The result: practices using BAM AI for services PA today can extend to drug PA with configuration, not implementation. No new vendor evaluation. No new integration project. No new staff training. The same agents, the same dashboard, the same workflow — just handling a broader scope of authorizations.

Preparing for the 2027 ePA Mandate: Action Steps for Healthcare Leaders

The CMS drug PA expansion hasn't been finalized yet — it's a proposed rule. But the direction is unmistakable, and practices that wait for finalization to start preparing will be 12-18 months behind. Here's the action sequence:

Step 1: Audit Your Current Drug PA Volume and Process

Most practices don't know their true drug PA burden because it's distributed across multiple staff members, tools, and channels. Count every drug PA your practice processes per week — including specialty pharmacy authorizations, formulary exceptions, step therapy overrides, and quantity limit appeals. The number is almost always higher than expected.

Step 2: Evaluate Infrastructure vs. Point Solutions

If you're already using AI for services PA, ask your vendor one question: "Can your system handle drug PA within the same platform?" If the answer involves a separate product, a new integration, or "we're building that," you have a point solution, not infrastructure.

Step 3: Consolidate Services and Drug PA Under One System

The CMS mandate will require unified ePA reporting. Practices running separate systems for services and drug PA will need to reconcile data across platforms for compliance reporting. A unified system eliminates this problem entirely.

Step 4: Build FHIR Readiness Now

The 2027 ePA mandate requires FHIR R4 API support. Whether you build, buy, or partner, ensure your PA automation vendor supports FHIR-based submission for both services and drugs. This isn't a nice-to-have — it's the compliance baseline.

Step 5: Deploy AI Before the Mandate, Not After

Practices that deployed AI for services PA before the 2027 mandate are already seeing ROI — reduced denial rates, faster approvals, lower staffing costs. The same logic applies to drug PA. Deploy now, learn on the current workflow, and you're positioned to absorb the expanded mandate seamlessly.

The $15 billion CMS projects in savings doesn't distribute itself evenly. It flows to organizations that automate first — capturing efficiency gains while competitors are still hiring PA coordinators and buying fax machines.

CMS is rebuilding the prior authorization operating model. The question isn't whether drug PA gets standardized — it's whether your practice automates it before the deadline or scrambles after.

CMS Is Coming for Drug Prior Auth Next — Is Your Practice Ready?