Medicare Enters a New Era With AI-Powered Prior Authorization
For decades, traditional Medicare stood apart from private insurance plans in one notable way: it largely operated without requiring prior authorization for most medical procedures. That approach gave physicians and patients considerable freedom, but it also left the system exposed to overuse, unnecessary treatments, and outright fraud. Now, the Centers for Medicare & Medicaid Services (CMS) is testing a fundamentally different approach — one powered by artificial intelligence.
In January 2025, CMS officially launched the WISeR Model, short for Wasteful and Inappropriate Service Reduction Model. The program is designed to reduce costs, ease administrative burdens on healthcare providers, and bring a new layer of oversight to a subset of procedures that have historically been vulnerable to misuse. For the first time in traditional Medicare's history, AI is being used to review prior authorization requests — and the early results are already drawing national attention.
What Is the WISeR Model and How Does It Work?
The WISeR Model is a CMS Innovation Center initiative that deploys AI-assisted prior authorization within traditional Medicare. Under the program, physicians in six participating states — Arizona, New Jersey, Ohio, Oklahoma, Texas, and Washington — must seek advance approval before performing a targeted set of medical procedures. These are procedures that CMS has identified as being at elevated risk for overuse or fraudulent billing.
Rather than routing every request through a slow manual review process, CMS has contracted with third-party technology companies to handle the initial evaluation using artificial intelligence. These AI systems analyze incoming prior authorization requests and issue preliminary decisions, often within the same business day. However, the model includes an important safeguard: a licensed clinician must review and sign off before any request is formally denied. This human oversight layer is intended to protect patients from inappropriate AI-driven rejections and to ensure medical decisions remain accountable to qualified professionals.
By combining the speed of AI with the judgment of licensed clinicians, the WISeR Model attempts to strike a balance between efficiency and patient safety — two goals that have often seemed at odds in healthcare administration.
Early Approval Rates: What the Data Shows in Texas
One of the most closely watched aspects of the WISeR pilot is its approval rate data, particularly from Texas, where Cohere Health serves as the contracted technology vendor. According to reporting by Washington Post Intelligence, approximately 62% of prior authorization requests in Texas receive approval on the first submission. After physician review, that number climbs to 84% — a meaningful jump that underscores the value of the appeals and review process built into the model.
Brian Covino, Chief Medical Officer at Cohere Health, noted that most approved requests receive a same-day response. This is a significant improvement over the administrative timelines that have long frustrated providers, who often wait days or even weeks for authorization decisions under traditional insurance processes. For patients who need timely access to care, faster turnaround times can have a direct impact on health outcomes.
The 62% first-pass approval rate has prompted some discussion about whether AI review is being calibrated appropriately — but proponents argue that the ability to achieve an 84% final approval rate after clinician review demonstrates the system's built-in flexibility and commitment to appropriate care delivery.
Why Traditional Medicare Needed a Prior Authorization Solution
Traditional Medicare's lack of prior authorization has long been considered both a feature and a flaw. On one hand, it reduced bureaucratic barriers and made it easier for seniors to access care quickly. On the other hand, it created conditions where certain high-cost or high-volume procedures could be performed without much scrutiny — opening the door to waste and fraud that ultimately raises costs for taxpayers and the Medicare program as a whole.
The WISeR Model directly addresses this vulnerability. By applying AI-driven review to a carefully selected list of procedures, CMS aims to identify inappropriate requests before services are rendered rather than attempting to recover payments after the fact. This proactive approach is considered more effective and less disruptive than post-payment audits, which often place significant burdens on providers and rarely recover a meaningful share of improper payments.
Key Goals of the WISeR Model
- Reduce wasteful spending: By flagging procedures that fall outside established clinical guidelines, the model targets a known source of unnecessary Medicare expenditures before payments are made.
- Combat fraud: AI review creates a consistent, data-driven check on requests that might otherwise slip through an overburdened manual review process.
- Ease administrative burden: Same-day responses and streamlined workflows are intended to reduce the paperwork and wait times that have long been a source of frustration for healthcare providers.
- Preserve care delivery: With a licensed clinician required to approve all denials, the model aims to ensure that legitimate medical needs are not blocked by automated systems.
What This Means for Physicians and Patients
For physicians practicing in the six pilot states, the WISeR Model introduces a new step in the workflow for covered procedures. While this adds a layer of administrative process, the promise of faster decisions — particularly same-day approvals — could ultimately make the experience less burdensome than prior authorization processes common in Medicare Advantage and commercial insurance plans.
For Medicare beneficiaries, the stakes are equally significant. Any prior authorization program carries the potential risk of delayed or denied care. The WISeR Model's design attempts to mitigate this by requiring human clinical review before denials are finalized and by targeting only a narrow set of procedures identified as high-risk for misuse. Seniors in the six participating states should be aware of the new requirements and communicate openly with their physicians about any potential delays or authorization needs.
A Pivotal Moment for AI in Federal Healthcare Programs
The WISeR Model represents a landmark moment in the evolution of AI within federal healthcare programs. Traditional Medicare covers tens of millions of Americans, and any changes to its administrative structure have broad implications for providers, patients, and the healthcare industry as a whole. If the pilot demonstrates that AI prior authorization can reduce costs and fraud without meaningfully disrupting care, CMS may have a strong case for expanding the program beyond the initial six states.
Conversely, if the model generates significant pushback from physicians or evidence of delayed care emerges, CMS will need to recalibrate both the technology and the procedures covered. Healthcare advocates and industry observers will be watching approval rates, denial patterns, and patient outcomes closely as the model matures.
What is clear is that the integration of AI into Medicare's administrative infrastructure is no longer a theoretical possibility — it is happening right now, in real clinical settings, affecting real patients and providers. The WISeR Model may well serve as the blueprint for how the United States healthcare system harnesses artificial intelligence to become more efficient, more equitable, and more resistant to the waste and fraud that cost the system billions of dollars each year.