U.S. SENATE JUL 21, 2026 · Updated July 28, 2026 · InsureAI Wire

Senate Blocks Effort to End AI Prior Authorization in Traditional Medicare

The Senate rejected the motion to proceed to S.J. Res. 198 on July 16, 2026, by 46 yeas to 50 nays, with four senators not voting. The resolution, a Congressional Review Act disapproval of the rule behind CMS’s Wasteful and Inappropriate Service Reduction (WISeR) Model, never reached a vote on its own merits. AI-assisted prior authorization therefore continues in traditional Medicare, even as congressional scrutiny of algorithmic coverage decisions intensifies.

The vote reached the floor at all because the Government Accountability Office decided in May, in B-337994, that the notice implementing WISeR is a rule for Congressional Review Act purposes, and that CMS never submitted it to Congress as one. That decision is what put a disapproval resolution on the expedited track. An identical earlier resolution, S.J. Res. 192, was discharged from committee on June 9 and has been on the Senate calendar ever since.

CMS announced WISeR in June 2025 and began it on January 1, 2026, in New Jersey, Ohio, Oklahoma, Texas, Arizona, and Washington. The agency’s model page describes a partnership with technology companies to test enhanced prior authorization for selected items and services. Participation is described as voluntary for providers, but with a catch: providers and suppliers in the six states must either submit prior authorization requests through the AI-assisted channel or face pre-payment medical review of their claims. It applies to traditional Medicare fee-for-service beneficiaries, not to Medicare Advantage plans.

The governance signal is distinct from the Senate probe of Medicare Advantage AI denials that surfaced the same week. That inquiry targets private insurers’ use of algorithms in Medicare Advantage coverage decisions. WISeR is a government-run demonstration using AI in traditional Medicare, and the Senate declined even to debate ending it on a party-line motion. Read narrowly, that is a procedural outcome rather than an endorsement. Read against the same week’s inquiry into private carriers, it still leaves algorithmic prior authorization advancing in the public program while the private one is asked to account for it.

Health insurers now operate against two different standards at once. Medicare Advantage plans face congressional demands for documented human review, an inventory of the predictive tools in use, and appeal records for AI-assisted denials. Traditional Medicare, through WISeR, is expanding AI use with fewer transparency requirements, though CMS has committed that a decision a service does not meet Medicare coverage requirements is made by a licensed clinician rather than by the technology. Carriers serving both markets have to hold their governance documentation to the stricter Medicare Advantage standard while tracking how CMS’s own AI deployment shapes what regulators end up treating as acceptable.

With no second vote scheduled, the next checkpoint is the independent, federally funded evaluation CMS has commissioned to inform whether the model can be expanded. The model runs in two three-year agreement periods through December 31, 2031, the first of which closes at the end of 2028. CMS has named what participants must report while it runs, including non-affirmations, decision times, and requests for re-review. A publication date for the evaluation itself is not on that list.

Share

Information aggregation and analysis, not legal advice. See our disclaimer.