AI-driven prior authorization: New ‘WISeR’ CMS model is right around the corner

Aug 22

As healthcare providers continue to explore the application of artificial intelligence (AI) in revenue cycle management, CMS is doing the same. Specifically, the agency recently announced it would test a new model for AI-driven prior authorization for items or services deemed wasteful or prone to fraud, including, but not limited to, electrical nerve stimulator implants, knee arthroscopy for knee osteoarthritis, and skin and tissue substitutes. It’s part of a larger push industry-wide to reduce burdensome prior authorization requirements. The CMS model Wasteful and Inappropriate Service Reduction (WISer) will leverage AI and other enhanced technologies (e.g., machine learning and algorithmic decision logic) to expedite the prior authorization process for these items and services with the goal of reducing inappropriate utilization, lowering spending in Original Medicare, expediting decision making, and easing provider administrative burden. Here’s what you can expect from CMS’ WISeR model for AI-driven prior authorization.

Important details of the WISeR model
Beginning on January 1, 2026, the WISeR model for AI-driven prior authorization will run for two three-year agreement periods, until December 31, 2031, and include four Medicare Administrative Contractor jurisdictions (MAC) and six specific states within those MAC jurisdictions, including New Jersey, Ohio, Oklahoma, Texas, Arizona, and Washington.

What’s unique about the WISeR model for AI-driven prior authorization is that it’s the first Innovation Center model in which technology innovators are the only model participants. Participants must be AI-driven prior authorization companies with expertise in providing recommendations on the medical necessity of coverage for payers using enhanced technology. One caveat related to the WISeR model? Clinicians with expertise in conducting medical reviews must validate each determination.

The WISer model is also the first model that incentivizes the use of cutting-edge tools to ensure that payment complies with Medicare documentation, coverage, and payment and coding rules. Applications were due July 25, and as part of the application process, these are just a few of the questions that CMS required AI-driven prior authorization technology companies to answer:

Describe your experience with health care-specific data, including experiences with and knowledge in applying prior authorizations, pre-payment reviews, and implementing coverage determinations set by health care plans and/or CMS.
Describe your clinical experience in medical reviews for coverage determinations and peer-to-peer reviews for resubmissions.
Describe your company/organization’s arrangement or staffing structure to access dedicated clinicians with relevant experience and expertise to conduct medical reviews to support clinically sound coverage determination decisions.
Describe your company’s integration with electronic health record platforms to deploy prior authorizations.
Describe how your company/organization’s approach, tools, and software will manage potential adverse effects of automation and AI.
Please provide any details related to approaches and quality assurance to utilization management decision making, complaints and grievances, claims billing, and claim dispute issues.

What physicians need to know about AI-driven prior authorization
Providers in selected regions will have the choice to submit a prior authorization request for the model’s selected items and services directly to the WISeR model participant or go through their MAC to do so. If a provider doesn’t submit a prior authorization for selected items and services, the claim would be subject to a pre-payment review, and the WISeR model participant or MAC may request additional documentation before paying the claim. The agency is also testing a gold card exemption for AI-driven prior authorization, whereby providers with a 90% approval (affirmation) rate during a performance review period of the WISer model could be exempted from future prior authorization requirements.

Some experts say while well-intentioned, the AI-driven prior authorization model has the potential for negative outcomes that include risk of care delays, increased administrative burden, and improper incentives for technology companies whose payments will be based on the number of prior authorization requests denied.

Five steps physicians can take now to prepare for AI-driven prior authorization
Here are five steps physicians can take now to prepare for the AI-driven prior authorization:

1. Know whether you’re located in a state and MAC jurisdiction impacted by the WISeR model. If you are subject to AI-driven prior authorization requirements, be sure to register for MAC updates so you can stay on top of important information.

2. Identify which service lines may be affected by AI-driven prior authorization and review corresponding CMS coverage criteria. Provide training on medical necessity and documentation requirements to avoid denials.

4. Establish workflows to ensure compliance with AI-driven prior authorization requests and/or pre-payment reviews.

5. Develop internal protocols for responding to denials and managing appeals related to AI-driven prior authorization.

6. Monitor AI-driven prior authorization affirmation rates to aim for gold-card eligibility.

Leverage the right technology
As medical practices prepare for AI-driven prior authorization inherent in the WISeR model, leveraging the right technology for streamlined, electronic workflows is critical. Learn how edgeMED can help.