Medicare's AI prior-authorization pilot drew a federal reprimand after delays and disputed denials

Medicare's WISeR pilot, launched January 1, 2026 in six states, uses AI to screen certain doctor-ordered procedures for prior authorization, with contractors paid a share of the spending their denials avert. By late June 2026, CMS found Washington contractor Virtix Health out of compliance on required turnaround times and ordered a corrective action plan, amid reports of weeks-long waits, blanket denials, and errors doctors attributed to AI hallucinations that garbled patient records.

Centers for Medicare and Medicaid Services (Virtix Health) · Incident Jun 22, 2026 · Indexed Jul 10, 2026 · 3 sources

Records by entity: Centers for Medicare and Medicaid

Contractors are paid a share of what their denials save, and doctors describe denials citing symptoms the records never contained.
What
Medicare's WISeR pilot, launched January 1, 2026 in six states, uses AI to screen certain doctor-ordered procedures for prior authorization, with contractors paid a share of the spending their denials avert.
Incident date
Jun 22, 2026
Who
Centers for Medicare and Medicaid Services (Virtix Health)
Failure mode
Policy Violation
AI surface
Agentic Workflow
Severity
High

What happened

The Wasteful and Inappropriate Service Reduction (WISeR) model launched January 1, 2026 in Arizona, New Jersey, Ohio, Oklahoma, Texas, and Washington, applying AI-assisted prior authorization to 13 traditional-Medicare services. Doctors and patients reported six-to-eight-week delays, blanket denials, and errors, including a denial citing numbness a radiologist had documented four times was absent, and a denial for the wrong region of the spine. Contractors are paid from the spending their denials avert, which critics say incentivizes denial regardless of merit, and roughly 80 percent of Medicare Advantage denials are overturned on appeal while few patients appeal. In late June 2026 CMS found Washington contractor Virtix Health out of compliance on the 72-hour turnaround requirement and ordered a corrective action plan, its first acknowledgment of the pilot's problems.

What broke inside the model

Failure path · mode profile · Policy Violation
  1. 01 · TriggerA prompt pushes against a deployment boundary.
  2. 02 · Model stepThe model produces the disallowed output.
  3. 03 · Control gapNo enforcement blocks it at generation time.
  4. 04 · FailureThe output crosses the policy line.
  5. 05 · ConsequenceA limit the business set is breached in public.

The output crosses a policy boundary the deployment had defined.

Vendors say humans make the final call, but clinicians describe AI playing a large role and producing denials that garble or fabricate details from patient records, a hallucination pattern applied to coverage decisions. The deeper design flaw is incentive alignment: an automated screening system whose operators profit from averted spending, run without transparency into denial and appeal rates, tilts toward denial and pushes the cost of error onto patients and doctors through delay and appeals.

Public visibilityHigh
Regulatory exposureActive
Customer impactClass-wide
Financial impactUnknown
Time to disclosureMonths
  1. PressFeds reprimand private company using AI to review WA Medicare claims over delayed processingkuow.org
  2. PressMedicare's AI push snarls patients and doctors in errors and delaysmedicalxpress.com
  3. PressCMS Slaps Its Own A.I. Contractor in Controversial WISeR Programhealthcareuncovered.substack.com
Permalinkhttps://failureindex.ai/failures/medicare-wiser-virtix-ai-prior-authorization-denials
CitationAI Failure Index. "Medicare's AI prior-authorization pilot drew a federal reprimand after delays and disputed denials" (FI-0712). Realm Labs. https://failureindex.ai/failures/medicare-wiser-virtix-ai-prior-authorization-denials (indexed Jul 10, 2026).
Share cardA branded image of this record for posts and slides.

Data fields CC-BY 4.0, prose citation permitted. Incident ID FI-0712. Full dataset at /data.

Note from Realm Labs, the Index steward

How Realm would have caught this

Controls for this failure mode
  • Prism
  • OmniGuard

Realm checks an AI-generated determination against the source record before it is issued, flagging a denial that cites facts the documentation does not support rather than letting a hallucinated rationale become a coverage decision. The runtime audit trail also exposes denial and override patterns that an opaque, incentive-driven pipeline would otherwise keep hidden from oversight.