EFF releases CMS records detailing delays and denials in Medicare WISeR pilot

The Electronic Frontier Foundation released approximately 1,000 pages of Centers for Medicare & Medicaid Services records on September 18, providing new details about delays, technical problems and provider complaints in Medicare’s AI-assisted WISeR prior-authorization pilot. The disclosure gives providers and patients a partial view of how the six-state program handled requests during its opening months.

EFF said the records show that two WISeR vendors denied 5,944 prior-authorization requests during the program’s first three months and that one request remained unanswered for 83 days. EFF corrected an earlier version of its post that misstated the denial total. The records available for review do not include a complete vendor-by-vendor denominator, appeal rate or reversal rate, so those figures do not establish a program-wide denial rate.

The records came from EFF’s Freedom of Information Act litigation against CMS. EFF’s case index lists participation agreements, a June 2025 investment plan, a reviewer guidebook, a data reporting guide, status-report emails, request and submission data, and provider feedback surveys. EFF says CMS is continuing to produce records.

CMS describes WISeR, the Wasteful and Inappropriate Service Reduction Model, as a six-year test covering selected Original Medicare items and services in Arizona, New Jersey, Ohio, Oklahoma, Texas and Washington. It began January 1, 2026, and is scheduled to run through December 31, 2031. The agency says the model uses technologies including artificial intelligence and machine learning together with human clinical review.

A non-affirmation means a request did not satisfy the applicable Medicare coverage, coding or payment requirements. CMS says technology cannot make that recommendation by itself: every non-affirmation must be reviewed by a clinician with relevant expertise. If the provider delivers the service and submits a claim after a non-affirmation, the Medicare Administrative Contractor denies the claim, but the existing Medicare appeals process remains available.

CMS says WISeR does not change Medicare coverage or payment policy. Instead, it moves review under existing coverage criteria earlier in the process for selected services. Providers can seek authorization through a WISeR participant or a Medicare Administrative Contractor, or deliver a selected service without prior authorization and face pre-payment medical review.

The agency’s 72-hour turnaround requirement applies to requests submitted to a participant portal. Requests routed through a Medicare Administrative Contractor can require additional processing, so the full path can take longer than the vendor’s review. CMS excludes inpatient-only services, emergency services and services for which a delay presents substantial risk. Providers may request expedited review when a delay could seriously jeopardize a patient’s life or health; CMS gives participants two days after they confirm that risk.

The released provider surveys contain allegations that delays and communication failures left patients in pain and forced surgery cancellations. Those accounts are provider complaints, not independently adjudicated findings that WISeR caused patient harm.

EFF also says the records show Innovaccer warned CMS about a month before launch that its software lacked full functionality and full testing. According to EFF’s account of the letter, Innovaccer planned to temporarily auto-affirm requests because CMS had not postponed the start date.

Separately, Representative Suzan DelBene’s office said CMS required Virtix to submit a corrective action plan for failing to meet the 72-hour review requirement. The available records do not provide enough data to assess the medical appropriateness of individual decisions.

The model’s payment method is another point of dispute. CMS says participants receive a performance-adjusted percentage of spending associated with care deemed wasteful or inappropriate. EFF concludes that paying participants from a share of averted spending creates a financial incentive to deny care. That characterization is EFF’s conclusion, not a CMS finding.

CMS says it audits participants, links accuracy findings to quality scores and payment adjustments, permits unlimited resubmissions, and may terminate participants with high inaccuracy. The agency also says a successful appeal can trigger withholding or recoupment of participant payment.

In a May 12 decision, the U.S. Government Accountability Office concluded that the WISeR notice was a rule subject to Congressional Review Act submission requirements and said the Department of Health and Human Services had not submitted the required report. HHS told GAO it considered the notice guidance and argued that the law did not apply.