AI is now reviewing Medicare claims. Here's what insurance professionals need to know.
The Centers for Medicare and Medicaid Services launched an artificial intelligence pilot program on January 1 that will review prior authorization requests for 6.4 million traditional Medicare beneficiaries across six states. The Wasteful and Inappropriate Service Reduction (WISeR) Model runs through 2031 and uses machine learning to evaluate 17 procedures in Arizona, New Jersey, Ohio, Oklahoma, Texas, and Washington.
The program targets an estimated $5.8 billion in Medicare spending on services with little or no clinical benefit. But early data shows the system creates significant problems for beneficiaries - and a critical role for professional advocates.
How the incentive structure creates bias
Four technology vendors administer WISeR: Cohere Health, Genzeon Corporation, Humata Health, and Innovaccer. Most operate under a "percentage-of-savings" payment model, where they receive a cut of the money they save through denials.
This structure creates built-in pressure to increase denial rates. No contract language can fully neutralize this incentive misalignment.
AI systems also produce disparate outcomes across populations. When training data is incomplete or historically skewed, the algorithms inherit those biases. Higher denial rates for marginalized groups often remain hidden in aggregate approval statistics.
Currently, no federal mandate requires insurers to disclose when an algorithm influenced a denial. This lack of transparency allows bias to operate without public oversight.
The denial numbers tell the story
Research from the American Medical Association shows AI-assisted reviews produce denial rates up to 16 times higher than human-only reviews. Early field reports from providers and patients in pilot states confirm this trend.
Yet most denials go unchallenged. Medicare Advantage insurers made 53 million prior authorization determinations in 2024, denying 7.7% of them. Only 5% of all denials - roughly one in nine - were ever appealed.
When beneficiaries do appeal, success rates depend heavily on service type:
- Hospital discharge decisions: ~80% overturn rate
- Durable medical equipment: 63.9% overturn rate
- Ambulance transport: 26.3% overturn rate
- Outpatient services: as low as 22.2% overturn rate
In 2024, only 7% of appealed Medicare Advantage denials were overturned. Most beneficiaries treat a machine's "no" as final because they don't understand their appeal rights.
Building your advocacy toolkit
Beneficiaries have 120 days from the denial notice to file an appeal. Appeals backed by clinical documentation, physician statements, treatment history, and peer-reviewed evidence succeed 82% of the time.
The appeals process has five levels:
- Level 1: Redetermination - Standard (60 days) or expedited (72 hours)
- Level 2: Reconsideration - Independent review entity with no plan affiliation
- Level 3: Administrative law judge hearing - Available when disputed amount exceeds $190
- Level 4: Medicare appeals council - Departmental escalation
- Level 5: Federal District Court - Final recourse when dispute exceeds $1,760
Overturn rates vary dramatically by insurer - sometimes by nearly a factor of two. As of March 31, CMS mandates all affected payers publicly post prior authorization metrics. Use these approval and denial rates during annual enrollment to inform plan recommendations.
The readiness review checklist
About 34% of insured Americans rank prior authorization as their biggest healthcare burden, above premiums, deductibles, and copays. The appeal burden is increasingly falling on Medicare Advantage enrollees, with appeals per enrollee nearly doubling since 2016.
Many beneficiaries in WISeR pilot states chose traditional Medicare specifically to escape prior authorization complexity. Your enrollment conversations need to expand beyond plan selection to include procedure planning.
Before a client schedules a procedure, conduct a WISeR readiness review:
- Identify "gold-card" providers - doctors who maintain 90%+ approval rates and are exempt from AI-driven reviews
- Know the 17 service categories most scrutinized under WISeR
- Build documentation checklists for high-risk procedures before claims are submitted
- Master the 72-hour expedited appeal window for urgent medical needs
The human element remains essential
AI is a decision-support tool, not a decision-maker. In a system optimized for savings over care accuracy, human oversight is the only thing standing between a beneficiary and a preventable medical crisis.
Your role has shifted from enrollment professional to healthcare navigator. The question CMS will face as it considers national expansion isn't whether algorithms will review claims - it's who will advocate for the person on the other side of the screen.
That person should be you.
Learn more about AI for Insurance or explore AI for Healthcare to build expertise in these systems.
Your membership also unlocks:
