{"resourceId":"cms-wiser-medicare-ai-prior-auth-vendor-flags-eff-foia-20261102","versions":[{"version":"external-1f69dc31e10ad7b52a28fc2be429b37bb6520c4a305dce946f684c81d2b59539","resource":{"id":"cms-wiser-medicare-ai-prior-auth-vendor-flags-eff-foia-20261102","title":"CMS WISeR Medicare AI prior-authorization pilot: FOIA records show four of six vendors flagged for delays, questionable denials and portal gaps","organization":"Centers for Medicare & Medicaid Services (WISeR model), reported by Nextgov/FCW","sector":"Public health coverage determination / prior authorization (traditional Medicare)","geography":"New Jersey, Ohio, Oklahoma, Texas, Arizona, and Washington, United States","publishedAt":"2026-11-02","publicationDate":"2026-11-02","eventDate":"2026-11-02","sourceName":"Nextgov/FCW","sourceLabel":"Federal technology trade press (FOIA-based reporting)","sourceUrl":"https://www.nextgov.com/artificial-intelligence/2026/11/medicares-ai-assisted-coverage-review-pilot-hit-early-snags-documents-show/416978/","evidenceClass":"independent-reporting","outcomeClass":"cautionary","topics":["governance-procurement","operating-model","data-security","knowledge-work"],"finding":"FOIA documents released by EFF and reported by Nextgov show CMS sent RFIs or corrective-action requests to four of six WISeR AI prior-authorization vendors early in 2026 over slow determinations, questionable or blank denials, missing electronic portals and weak provider communication. Named vendors include Genzeon (New Jersey), Virtix Health (Washington), Zyter (Arizona) and Innovaccer (Ohio). CMS later said no CAPs are active and vendors are compliant, with a six-month performance report forthcoming. CMS deputy administrator Kimberly Brandt publicly acknowledged many things about WISeR are not working; House Appropriations advanced an amendment to suspend WISeR funding.","sledRelevance":"A large federal AI-assisted coverage-review experiment in six states with documented early vendor noncompliance and congressional scrutiny—directly instructive for state Medicaid and public-hospital prior-auth AI programs considering similar vendors and human-in-the-loop claims.","evidence":"Nextgov synthesis of EFF FOIA RFIs/CAPs, vendor response letters, CMS spokesperson comments and public remarks by CMS leadership. Early-period performance; later improvements claimed by vendors/CMS without independent audit in this article.","architectureImplications":"Interpretation: vendor AI plus licensed clinician review for selected Medicare services, with provider portals and determination-letter workflows; Zyter noted as using Palantir infrastructure in related EFF material cited by secondary coverage. Technical details of models are not in the Nextgov piece.","governanceImplications":"Rushed six-month launch, active CAP/RFI oversight, congressional funding threat and limited public transparency on whether human reviewers meaningfully override AI recommendations.","securityPrivacyImplications":"Coverage requests include clinical detail; article focuses on operational failures more than cybersecurity, but portal readiness failures are themselves access-control risks.","caveats":"Federal Medicare WISeR, not Medicaid; documents describe early-2026 problems and CMS asserts current compliance. Choose Nextgov over MedCity/EFF as the Nov. 2 independent-reporting synthesis.","streamIds":["health-human-services"],"roles":{"sales":"Sales framing: WISeR is a cautionary proof that AI prior-auth sold on savings will be judged on turnaround, denial quality and portal readiness under FOIA and congressional heat. For Medicaid and public-hospital buyers, lead with readiness criteria, published SLAs and independent accuracy audits—not with denial rates. Do not claim WISeR is fixed; CMS itself says many things are not working.","engineering":"Engineering framing: prior-auth AI needs a production-grade provider portal on day one, deterministic determination-letter generation, duplicate-tracking controls and latency SLOs, plus an auditable clinician override path. Rushed launches without those controls produced duplicate tracking numbers, blank letters and multi-day stalls in the FOIA record. Design for public auditability of recommendation versus final decision.","delivery":"Delivery framing: stage go-live behind portal acceptance tests, 72-hour determination SLAs and denial-quality sampling with clinicians empowered to overturn AI recommendations. Plan provider education so questionable denials are not blamed solely on user error, and prepare operations for FOIA and legislative document requests from day one."},"roleDetails":{"sales":{"customerProblem":"Coverage programs want AI to cut wasteful spend without delaying care or generating indefensible denials.","stakeholders":["Medicaid/Medicare coverage leads","Hospital association and specialty providers","Legislative oversight staff"],"discoveryQuestions":["What go-live readiness gates exist for the provider portal?","How are denial-accuracy samples reported publicly?","Can clinicians override without quota pressure?"],"valueHypothesis":"Bounded: AI-assisted review may speed routine affirms if portals and SLAs work; early WISeR evidence shows the opposite risk when rushed.","potentialEngagement":"Prior-auth AI readiness and denial-quality assessment for a state coverage program.","unsupportedClaims":["That WISeR currently meets all SLAs","That human review fully prevents improper denials","Net savings after access harms"]},"engineering":{"fit":"AI-assisted prior authorization with mandatory licensed-clinician review for non-affirmations.","architectureIntegration":"Provider portal, case tracking IDs, AI recommendation, clinician review, determination letters and claims integration.","prerequisites":["Portal ready before launch","Turnaround SLO monitoring","Denial rationale templates tied to coverage criteria"],"constraints":["Compressed implementation timelines","Vendor incentive to maximize prevented spend","Opaque model features"],"security":"Secure provider authentication, PHI minimization in vendor environments, and immutable determination audit logs.","proofOfValue":["Share of determinations within SLA","Clinician override rate","Provider resubmission/affirmation rates after education"]},"delivery":{"work":"Phased service list, portal UAT, clinician reviewer staffing, provider education, public performance reporting.","dependencies":["CMS/state model rules","Vendor portal delivery","Clinical reviewer capacity"],"ownership":"Coverage operations with compliance and IT.","skillsAdoption":"Train clinicians to challenge AI recommendations and document independent rationales.","governanceCheckpoints":["Pre-launch readiness review","Weekly CAP-style exception review","Public six-month performance report"],"acceptanceCriteria":["Proposed: zero go-live without functioning electronic submission","Proposed: determination SLA met for ≥95% of cases"],"risks":["Care delays and patient harm","Congressional or CMS suspension","Provider abrasion and resubmission churn"]}},"retrievedAt":"2026-11-05T22:09:47-06:00","enrichedAt":"2026-11-05T22:16:10-06:00","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Provider burden and patient delays (e.g., reported bedside distress waiting on kyphoplasty auth) show workforce and access harms when AI prior-auth underperforms.","procurementImplications":"Payment structure sharing savings from prevented 'wasteful' care creates denial-incentive concerns; readiness gates and enforceable SLAs should precede go-live.","operatingModelImplications":"Human-in-the-loop non-affirmation review is required by CMS FAQ, yet early letters suggest turnaround and denial-quality failures—operating model depends on reviewer capacity and authority, not slogans.","sourceVerification":{"openedUrl":"https://www.nextgov.com/artificial-intelligence/2026/11/medicares-ai-assisted-coverage-review-pilot-hit-early-snags-documents-show/416978/","referenceExcerpt":"There are a lot of things about WISeR that are not working,","promptVersion":"sled-research-v4.0 HHS nightly brief","model":null,"basis":"agent-reported inspection"}}}]}