If you are enrolled in a Medicare Advantage plan, you have almost certainly encountered prior authorization — the requirement that your insurer approve certain treatments, procedures, or medications before you can receive them. What most beneficiaries do not know is that many of the largest Medicare Advantage insurers have been quietly delegating those approval decisions, at least in part, to an artificial intelligence system called WISeR. Internal records reviewed by patient advocacy organizations, including the Medicare Rights Center, are now raising serious alarms about whether that AI is wrongly denying care that seniors legitimately need — and whether the humans nominally overseeing the system are doing enough to catch its mistakes.
WISeR stands for Widely Integrated Systematic Evidence Review. It is an AI-driven clinical decision-support tool that evaluates prior authorization requests by comparing a patient's clinical profile against a database of medical criteria and flagging requests that fall outside what the algorithm considers standard care. In theory, this speeds up the review process and reduces administrative burden on insurers. In practice, records suggest the system may be applying those criteria too rigidly — denying care that a human physician reviewing the same case would likely approve. The tool is not a fringe product used by small regional plans; it has been adopted by some of the largest Medicare Advantage insurers in the country, meaning its decisions touch millions of beneficiaries.
The fundamental problem with AI systems like WISeR is that they are trained on population-level data. They identify patterns across millions of patients and make probabilistic judgments about what care is typical for a given diagnosis. But medicine is not always probabilistic, and Medicare beneficiaries are rarely textbook cases. A 74-year-old woman recovering from a hip replacement may need an extra week of inpatient rehabilitation that falls outside what the algorithm expects for her age group and diagnosis code. A man managing heart failure, Type 2 diabetes, and chronic kidney disease simultaneously may require a specialist referral that does not fit neatly into the AI's decision tree. When the algorithm flags these cases as outside the norm, the result is a denial letter — even when the treating physician has clearly documented why the care is necessary for that specific patient.
According to CMS.gov data, Medicare Advantage plans collectively submitted more than 46 million prior authorization requests in a recent reporting year. The denial rate for those requests has become a growing concern for federal regulators, and CMS data from its Medicare Advantage and Part D Contract and Enrollment data files shows that the number of prior authorization denials subsequently overturned on appeal has been rising year over year. That pattern is significant: when a denial is overturned on appeal, it means the initial decision was wrong. When an AI system is making or heavily influencing those initial decisions at scale, even a modest error rate translates into tens of thousands of seniors being wrongly denied care — and only a fraction of those seniors will know to appeal.
The Medicare Rights Center has been tracking complaints related to AI-assisted prior authorization denials and found a troubling pattern: denial notices are frequently issued with boilerplate language that does not explain why the specific patient's circumstances were insufficient to meet medical necessity criteria. Federal law requires that denial notices include the specific clinical criteria used to make the decision, but advocates say that requirement is routinely not met in a meaningful way when AI systems generate the denial language. This matters practically because if you do not know exactly what the algorithm flagged as a problem, it is much harder for you and your doctor to craft an effective appeal.
Data Snapshot: According to CMS.gov data for 2025, there were more than 4,800 Medicare Advantage plans available nationally across all plan types, including Health Maintenance Organizations, Preferred Provider Organizations, and Special Needs Plans. CMS star ratings data for 2025 shows that fewer than 30 percent of Medicare Advantage contracts earned four stars or higher — the threshold CMS uses to identify plans with strong performance on member experience measures, which include appeals and complaints. Beneficiaries enrolled in plans rated below four stars are statistically more likely to encounter administrative friction, including prior authorization disputes. The average Medicare Part B premium in 2025 is $185.00 per month, a figure that underscores how much financial skin beneficiaries already have in the game before they pay a single plan premium or cost-share.
If you have received a prior authorization denial from your Medicare Advantage plan, the most important thing to understand is that you have a legally protected, multi-level appeals process — and it has real teeth. The first step is to request a redetermination from your plan, which must be completed within 60 days for standard appeals. If your health is at immediate risk, you can request an expedited appeal, which your plan must resolve within 72 hours — not business days, but calendar hours. If the plan upholds the denial, you escalate to an Independent Review Entity, a third-party organization contracted by CMS to review disputes independently of the insurer. After that, you can request a hearing before an Administrative Law Judge if the amount in controversy meets the threshold (currently $180 or more for 2025). You can then appeal to the Medicare Appeals Council and, ultimately, to federal district court. At each level, your plan must meet a higher burden to justify the denial, and studies have consistently shown that beneficiaries who appeal are successful a meaningful portion of the time.
Your treating physician is your most powerful ally in this process. When fighting an AI-generated denial, ask your doctor to write a detailed letter of medical necessity that goes well beyond standard diagnosis codes. The letter should explain, in plain clinical language, why the specific treatment is necessary for you as an individual — not for a hypothetical average patient with your diagnosis. It should reference your specific test results, functional limitations, prior treatment history, and why alternatives the plan might suggest are not appropriate for your particular case. This kind of individualized documentation is precisely what an AI system cannot generate on its own, and it is often what tips the scales in an appeal. Some physicians' offices have staff dedicated to prior authorization disputes; ask whether yours does.
CMS issued guidance in 2024 making clear that Medicare Advantage plans cannot use AI or algorithmic tools as the sole basis for a coverage denial — a qualified human clinical reviewer must be involved in the decision. However, enforcement of that requirement has been inconsistent. Patient advocates argue that in practice, AI recommendations are being rubber-stamped by human reviewers who lack the time or the complete clinical picture to meaningfully override the algorithm. Legislation has been introduced in Congress to require greater transparency about how AI tools are used in coverage decisions and to establish clearer accountability standards, though as of this writing those bills have not been enacted into law. CMS has indicated it is monitoring plan compliance with the 2024 guidance and that plans found to be in violation face potential sanctions, including civil monetary penalties.
For beneficiaries who want to take proactive steps before a denial occurs, there are several concrete actions worth taking now. Before you need a prior authorization, ask your doctor's office about their experience with your specific plan's approval process for the treatments you are likely to need. Practices that deal frequently with a particular insurer often know which services are routinely flagged and can preemptively document medical necessity more thoroughly in the initial request. Keep copies of all correspondence with your plan, including denial letters, appeal submissions, and any phone call records with dates and representative names. If you receive a denial, do not wait — the clock starts ticking immediately, and missing appeal deadlines can forfeit your rights entirely.
Contact your local State Health Insurance Assistance Program, known as SHIP, for free one-on-one counseling. SHIP counselors are trained specifically to help Medicare beneficiaries navigate coverage disputes, appeals, and plan comparisons at no cost to you. You can find your state's SHIP program at shiphelp.org. The Medicare Rights Center also operates a national helpline at 800-333-4114 and can provide guidance tailored to your specific situation. Both resources are independent of insurance companies and have no financial stake in the outcome of your dispute.
The Annual Enrollment Period runs from October 15 through December 7 each year, and during that window you can switch to a different Medicare Advantage plan or return to Original Medicare paired with a standalone Part D prescription drug plan. If you have experienced repeated prior authorization problems with your current plan, that is a meaningful signal worth acting on. When comparing Medicare Advantage plans during open enrollment, look specifically at the plan's CMS star rating — plans rated four stars or higher have generally demonstrated better performance on member appeals and complaints — and review the plan's prior authorization requirements for the specific services you use most. Medicare.gov's Plan Finder tool allows you to compare plans side by side on cost, coverage, and quality ratings.
If you believe you have been harmed by an inappropriate AI-driven denial, you can file a formal complaint through several channels. Medicare.gov's complaint system routes your grievance directly to CMS. The Medicare Rights Center can help you document and escalate your complaint. Your SHIP counselor can assist you in framing the complaint in terms that are most likely to prompt regulatory attention. These complaints matter beyond your individual case — CMS uses complaint and appeals data when evaluating plan performance, determining audit targets, and deciding whether to sanction insurers. The more beneficiaries report problems, the more pressure regulators have to enforce existing rules and develop stronger guardrails around AI use in coverage decisions.
The accountability gap at the heart of this issue is worth naming plainly. When a human physician denies a prior authorization request, there is a named professional who made a clinical judgment and can be questioned about it. When an AI system generates a denial, accountability becomes diffuse — the insurer points to the algorithm, the algorithm's developer points to the training data, and the patient is left holding a denial letter with no clear explanation of what went wrong or who is responsible. That gap is precisely what patient advocates are pushing regulators and lawmakers to close. Until stronger, consistently enforced rules are in place, the burden falls on individual beneficiaries to know their rights, document their cases thoroughly, and use every level of the appeals process available to them. The system is designed to be navigable — but only if you know it exists.