Federal health officials are testing artificial intelligence to screen Medicare claims for services they consider vulnerable to overuse, promising savings but raising alarms among doctors, patient advocates, and lawmakers about access to necessary care. The demonstration—called WISeR—combines machine learning with human review and is being piloted in six states through December 2031.
How the WISeR model works
The Centers for Medicare and Medicaid Services (CMS) started WISeR, short for Wasteful and Inappropriate Service Reduction Model, to identify procedures and products it views as prone to overuse, fraud, or abuse. The pilot targets items such as skin and tissue substitutes, certain electrical nerve stimulator implants, and knee arthroscopy for osteoarthritis. CMS says the model will pair automated tools with clinical review to ensure appropriate payments and reduce unnecessary procedures.
WISeR represents a significant shift: prior authorization has long been a common feature of Medicare Advantage plans but has been rarely used in original Medicare. By deploying AI to triage claims that may require extra scrutiny, CMS aims to catch potentially wasteful spending earlier in the payment process.
Clinical community and patient advocates push back
Physicians and advocates worry the new approach could produce wrongful denials and delayed care. A 2025 American Medical Association survey found that 61 percent of doctors fear AI will worsen denials of treatments they consider medically necessary. Critics argue that automating prior authorization risks turning a tool meant to curb unnecessary care into a barrier to appropriate treatment.
Health policy analyst Camm Epstein told Undark, “AI should be used to make appropriate care easier to approve, not necessary care easier to deny.” Advocates also point to reporting that the WISeR pilot has already been associated with care delays and denials in each of the six states where it is running.
For patients, prior authorization is frequently experienced as a burdensome obstacle. A Commonwealth Fund survey released recently found roughly one in five working-age adults with private insurance reported that they or a family member were denied coverage for physician-recommended care in 2025. Of those denied, 41 percent said the denial delayed care and more than a quarter reported their condition worsened as a result.
Evidence of wrongful denials in Medicare Advantage
Concerns are amplified by prior findings about Medicare Advantage plans. An HHS Office of Inspector General memo in 2022 cited instances—more than one in 10—where beneficiaries were denied access to services despite apparently meeting coverage rules. While plans overturned a large share of denials on appeal in 2024—about 81 percent—those reversals often come only after additional delay and administrative burden for patients and providers.
Those reversal rates underscore both the potential for error in automated or administrative denials and the heavy effort required to resolve them.
Incentives and transparency questions
Another flashpoint is how vendors implementing WISeR are paid. CMS says participating vendors receive a portion of what it calls “averted expenditures,” which could translate into revenue tied to reductions in services. That arrangement has prompted criticism that financial incentives could encourage denials rather than careful clinical judgment. Several lawmakers have proposed measures to block funding for WISeR, citing risks to patient access.
Industry groups have attempted to reassure stakeholders. In responses to a recent survey, all participating health plans stated they do not use AI or algorithms without clinician review to deny requests involving medical necessity or clinical considerations. Insurers also pledged to provide more transparency about the clinical reasoning behind denials. Still, advocates say such commitments do not fully address the potential for automated systems to generate inappropriate denials or create new administrative burdens for clinicians.
Regulatory context and competing priorities
The WISeR pilot comes amid broader federal efforts to reform prior authorization. A rule issued during the previous administration set strict timelines for government-run plans—72 hours for urgent requests and seven days for non-urgent requests—and those requirements took effect on Jan. 1 for many public plans. Separately, private insurers have pledged to standardize electronic prior authorization requests by 2027 and to reduce the volume of services subject to prior authorization by 2026 for certain common procedures.
CMS leaders have signaled a dual approach: expand scrutiny in original Medicare while pressing private plans to streamline and reduce prior authorization burdens. CMS Administrator Mehmet Oz warned insurers they must ease prior authorization or face federal regulation: “If you don’t do it yourselves, then we’re going to do it for you,” he said.
Industry data indicate some movement: a survey found prior authorization requests fell about 11 percent between June 2025 and April 2026. But it is not yet clear whether the rate of denials or the incidence of harmful delays has changed.
Why AI may not be a simple fix
Proponents say AI can rapidly sort claims to expedite clearly allowable approvals, potentially reducing delays for straightforward cases. Critics counter that automating parts of a deeply flawed system risks magnifying harms when algorithms are used to gate access to care. As Jared Dashevsky, a physician and founder of Healthcare Huddle, put it: while AI could “eliminate barriers, reduce administrative waste, give us more time with patients,” current developments look more like an “arms race to deny faster and appeal faster.”
As the WISeR demonstration proceeds, stakeholders will be watching for evidence that AI-enabled review both preserves patient access to necessary care and reduces true waste, not merely shift administrative burdens or create perverse incentives to deny services.
Source: Ars Technica AI