AI Prior Auth Pilot Delays Care for 6.4M Medicare Patients

The use of artificial intelligence to gatekeep medical care has entered a contentious new phase with the WISeR pilot. As a first-of-its-kind deployment of AI for prior authorization in Original Medicare, the system is already generating a backlash over weeks-long waits and questionable denials. For AI developers and health tech leaders, the pilot is a high-stakes proving ground—success could spur widespread AI adoption in clinical administration, but failure may set back trust in algorithmic healthcare decisions for years.

In a historic shift for Original Medicare, the Centers for Medicare & Medicaid Services (CMS) launched the Wasteful and Inappropriate Service Reduction (WISeR) pilot in January 2026, introducing prior authorization—and artificial intelligence—to millions of beneficiaries who never before faced such hurdles. For decades, traditional Medicare largely sidestepped the prior authorization burdens that have angered patients and providers in Medicare Advantage plans. The WISeR model, however, now requires pre-approval for a slate of common procedures across six states, including epidural steroid injections, treatments for spinal stenosis and knee osteoarthritis, and nerve stimulation for tremors and Parkinson’s disease. With approximately 6.4 million beneficiaries affected in New Jersey, Ohio, Oklahoma, Texas, Arizona, and Washington, the pilot is a six-year experiment designed to slash “crushing fraud, waste, and abuse,” in the words of CMS Administrator Mehmet Oz. The agency is contracting with private companies that use “enhanced technologies, including artificial intelligence” to adjudicate requests—a move intended to modernize oversight but one that immediately ignited controversy.

“

A report from Senator Maria Cantwell (D-Wash.) reveals that seniors in her state are now waiting up to three weeks for authorization decisions that once took days, with procedure backlogs stretching even longer.

Barely six months in, the rocky start that critics predicted is materializing. A report from Senator Maria Cantwell (D-Wash.) reveals that seniors in her state are now waiting up to three weeks for authorization decisions that once took days, with procedure backlogs stretching even longer. “For Washington’s seniors, WISeR is overriding doctors to delay care and deny treatment,” she stated, echoing the very frustrations that have plagued Medicare Advantage for years. The KFF health policy organization has long cautioned that prior authorization can lead to administrative snarls, patient delays, and denials of medically necessary services—and those predictions are now being borne out in the traditional program. These delays are not merely inconvenient; they risk progression of degenerative conditions and unnecessary suffering for patients with chronic pain or neurological disorders.

The AI component adds a layer of opacity and unease. While CMS has framed the use of technology as a tool for efficiency, providers and patients have no visibility into how the algorithms make decisions. The private contractors remain unnamed, and there is no publicly available information on the models’ accuracy, bias, or error rates. For a pilot that could pave the way for a nationwide expansion, the lack of transparency raises serious governance questions. AI-driven healthcare decisions are already under intense scrutiny; earlier setbacks like the IBM Watson Health debacle loom large, and the WISeR pilot risks reinforcing skepticism if it cannot demonstrate fairness and reliability.

The stakes extend beyond patient care. A botched rollout could chill congressional appetite for further AI integration in Medicare, hindering automation that could genuinely reduce fraud and administrative waste if implemented thoughtfully. Conversely, if CMS can iron out the early kinks—improving review speeds, publishing algorithm performance data, and establishing clinician appeal pathways—the pilot might eventually serve as a case study for responsible AI in public health insurance. The six-year timeline gives the agency room to iterate, but the political pressure from lawmakers like Cantwell may accelerate demands for a pause or redesign.