HEADER: A new Senate resolution seeks to overturn CMS’s WISeR pilot. This is the CMMI Medicare model that applies AI-supported prior authorization to selected services considered vulnerable to waste or inappropriate use.
- I wrote about some glitches in WISeR on May 11.
Senate Democrats, led by Ron Wyden, Maria Cantwell, Richard Blumenthal, and Kirsten Gillibrand, argue that the model risks importing the burdens of private-plan prior authorization into traditional Medicare, where such controls have historically been uncommon.
CMS describes WISeR as a targeted fraud-and-waste initiative focused on services such as skin and tissue substitutes and epidural steroid injections (in its Version 01, no genomic tests.) Critics counter that prior authorization can delay care, increase administrative burden, and place algorithmic tools between physicians and beneficiaries. The political opening came after GAO concluded that WISeR is agency rulemaking subject to the Congressional Review Act, despite HHS’s earlier position that the notice was merely guidance. That decision gives Congress a procedural path to force a vote on repealing the model.
 |
| Falls under CRA Congressional Review Act |
###
See the resolution here:
https://www.finance.senate.gov/imo/media/doc/052026_wiser_cra_resolution.pdf
See May 20 press by Wyden:
See an 18-page, May 12 legal memorandum by GAO on the procedural status of WISeR.
https://www.gao.gov/assets/890/885713.pdf
###
AI CORNER
###
See Chat GPT summarize the GAO memorandum.
GAO’s May 12, 2026 decision is narrow but important: CMS’s WISeR notice is a “rule” under the Congressional Review Act, so CMS should have submitted it to Congress and GAO before it took effect. GAO does not decide whether WISeR is good policy, whether AI prior authorization is lawful, or whether CMS can ever test such a model. The decision is about process and congressional review.
GAO’s reasoning is straightforward. The WISeR notice was an official CMS statement, published in the Federal Register, with future effect, and it created new procedures for selected Original Medicare services in six states. GAO rejected HHS’s argument that the notice was merely guidance; GAO has previously held that guidance documents can still be CRA “rules” when they implement or prescribe policy. (GAO)
The key substantive point is that WISeR changes who effectively makes payment-related decisions. Although MACs still formally approve or deny claims, GAO says they are bound by determinations made by WISeR model participants. Those participants use advanced technology, including AI or machine learning, and are compensated based on reducing Medicare expenditures. GAO found that this could affect whether claims are approved or denied.
GAO also rejected the idea that WISeR is voluntary in a meaningful sense for providers. Third-party contractors may choose to participate, but providers in the six states must use the WISeR procedures for selected services: either seek prior authorization or undergo pre-payment review by a model participant. GAO concluded that this substantially affects providers and beneficiaries, so the CRA’s procedural-rule exception does not apply.
Bottom line: GAO held that WISeR is not just an internal CMS experiment or informal guidance. It is a rule with real effects on Original Medicare claims processing, and therefore subject to Congressional Review Act submission requirements before taking effect.
###
###
Chat GPT pointed out the improved typography and design from GAO.
Hiding in plain sight is what appears to be a major redesign and new typography stylebook for GAO reports. The May 12 WISeR decision does not look like the old federal-document stereotype: dense text, cramped margins, weak hierarchy, and the faint ghost of a photocopier. Instead, the page has a consciously designed and quite up to date look. The GAO masthead is restrained and confident; the opening “Decision” heading functions as a true display element; and the metadata fields — “Matter of,” “File,” and “Date” — are organized on a clean grid rather than simply dumped into the page.
The typography is also notably more contemporary. The document appears to rely on a distinctive, modern sans-serif face, with strong weight contrast for headings and labels, generous leading, and enough white space to let the page breathe. This is not flashy design. It is better than that: both competent and appealing. It communicates authority, accessibility, and procedural seriousness without looking like a stylesheet that was embalmed in the WordPerfect era.
