AI Is Denying Your Health Insurance Claims in Seconds — What Every Policyholder Must Know in 2026
Photo by Kelly Sikkema on Unsplash
- Cigna's PXDX AI algorithm denied over 300,000 health insurance claims in just two months in 2022, spending an average of 1.2 seconds reviewing each one — effectively zero human review.
- UnitedHealthcare's nH Predict AI reportedly had a 90% error rate, yet was used to override doctors' recommendations for elderly Medicare Advantage patients, triggering a federal class action lawsuit.
- A 2024 U.S. Senate report found AI tools produced claims denial rates up to 16 times higher than those made by human reviewers at the same insurers.
- Fewer than 1% of denied claims are ever appealed — but 80% of Cigna Medicare Advantage AI denials that were appealed were ultimately overturned, suggesting widespread over-denial.
What Happened
If you've ever had a health insurance claim denied and wondered whether a real person actually reviewed your case, you now have good reason to ask. Over the past several years, major health insurers have quietly deployed artificial intelligence algorithms to automate the claims management process — and the results have triggered lawsuits, congressional investigations, and alarm from patients, physicians, and regulators across the country.
Here is what the evidence shows. In 2022, Cigna's PXDX AI system denied more than 300,000 claims in just two months, with an average review time of 1.2 seconds per claim. That is not a typo — 1.2 seconds, which leaves virtually no room for any human review of your medical records or your doctor's recommendations. Around the same period, UnitedHealthcare deployed an AI tool called nH Predict through its subsidiary NaviHealth to make coverage decisions for patients needing post-acute care (meaning care in a skilled nursing facility or at home after a hospital stay). The outcome was dramatic: its post-acute services denial rate climbed from 8.7% in 2019 to 22.7% by 2022, and its skilled nursing facility denial rate alone jumped ninefold — from 1.4% to 12.6%.
In 2024, the U.S. Senate Permanent Subcommittee on Investigations released a formal report criticizing UnitedHealthcare, Humana, and CVS for using algorithmic tools to sharply inflate claims denials for Medicare Advantage beneficiaries. Then on March 31, 2025, a U.S. District Court for the Eastern District of California allowed a class action lawsuit against Cigna over its automated PXDX system to proceed — a ruling that sets a significant legal precedent for how courts may treat AI-driven claims management going forward.
Photo by Vitaly Gariev on Unsplash
Why It Matters for Your Coverage
Building on those facts, it is worth stepping back to understand exactly why this trend is so consequential for ordinary policyholders.
Think of it this way: you go to the doctor, your physician reviews your history and determines you need a specific treatment or an extended stay in a care facility. Then, in the span of a heartbeat, a computer program overrides that recommendation — without any doctor on the insurer's side ever reading your file. That is precisely what a federal class action lawsuit alleges happened with UnitedHealthcare's nH Predict algorithm. The suit claims the model had a known 90% error rate, yet was used anyway to deny coverage for elderly Medicare Advantage patients who needed ongoing care after hospitalization.
This has direct consequences for your policy coverage. Health insurance is built on a core promise: it will pay for medically necessary care as defined by your plan. But when an AI system performs that medical necessity determination (the process of deciding whether a treatment is appropriate and covered under your plan) in 1.2 seconds, the risk assessment that was once done by licensed clinicians reviewing actual patient records is effectively handed to a statistical model trained on historical claims data.
The numbers reveal just how widespread the problem has become. According to the Experian State of Claims Report 2025, 41% of healthcare providers reported that more than 10% of their claims are now denied — up sharply from 30% in 2022 and 38% in 2024. A 2025 National Association of Insurance Commissioners (NAIC) survey found that 71% of health insurers are now using AI for utilization management (the insurer's process of deciding in advance whether a treatment or service is medically necessary before they will pay for it).
Washington State Insurance Commissioner Mike Kreidler called the practice of using AI to routinely deny claims "abhorrent," saying it was being done "to enhance the bottom line" rather than to genuinely assess medical necessity. The 2024 Senate committee report backed that view, finding that AI-generated denial rates were up to 16 times higher than what human reviewers at the same companies would have produced — a gap that speaks to the importance of doing a careful insurance comparison whenever you are evaluating health plan options.
And perhaps most revealing: fewer than 1% of patients ever appeal a denied claim, and of those who do, patients lose more than half. Yet when Cigna's Medicare Advantage AI denials were appealed, roughly 80% were overturned. As Rep. Cathy McMorris Rodgers pointed out, about one in five Cigna Medicare Advantage denials are even contested — meaning the overwhelming majority of potentially wrongful denials are simply accepted by patients who don't know they can fight back. Researchers publishing in JAMA Health Forum warned that this represents a fundamental shift in how medical necessity is determined, raising serious concerns about both due process and patient safety. For everyday policyholders, claims management has quietly become a one-sided contest — and most consumers don't realize they are already losing.
Photo by Trnava University on Unsplash
The AI Angle
The tools at the center of this controversy — Cigna's PXDX system and UnitedHealthcare's nH Predict — represent a broader wave of insurtech (insurance technology) automation reshaping the industry. These machine learning platforms are trained on vast historical claims datasets to predict whether a given treatment, procedure, or care stay falls within what the insurer considers statistically normal. Claims that deviate from the model's predictions are flagged or denied automatically.
Insurers market these tools as efficiency boosters and fraud-prevention systems, promising faster claims management and insurance savings through reduced administrative overhead. In theory, AI can legitimately accelerate routine approvals and catch clear-cut billing fraud. But the data makes clear that when the same tools are applied to prior authorization (the process requiring insurers to pre-approve certain treatments before care is delivered) and coverage denials — especially for complex or vulnerable patient populations — the risk assessment failures can be severe and widespread.
The legal system is now responding. The March 2025 ruling allowing the Cigna class action to advance signals that courts are prepared to hold insurers accountable not just for their human decisions, but for the algorithmic systems they deploy. AI-driven claims automation will remain one of the most contested areas in health insurance well into 2026 and beyond.
What Should You Do? 3 Action Steps
The evidence is unambiguous: most patients never appeal, but those who do frequently win. If your claim is denied, immediately request a written explanation stating the specific reason and the clinical criteria used. Then file a formal internal appeal within the timeframe your plan requires (usually 60 to 180 days). If that fails, you have a federal right to an independent external review by a third-party organization with no ties to your insurer. The 80% overturn rate on Cigna Medicare Advantage AI denials proves that persistence pays off. A licensed insurance agent or a patient advocate can help you understand your policy coverage rights and build the strongest possible appeal.
Not every insurer uses AI the same way, and denial rates vary significantly between plans. When shopping for health insurance — especially Medicare Advantage — do a thorough insurance comparison that goes beyond the monthly premium. The Centers for Medicare and Medicaid Services (CMS) publishes prior authorization denial rate data for Medicare Advantage plans, and consumer advocates recommend checking your state insurance commissioner's complaint records as well. Choosing a plan with more transparent claims management practices and lower historical denial rates can translate into meaningful insurance savings when you actually need care — because a lower premium means little if your claims are routinely blocked by an algorithm.
Keep detailed records of every interaction with your insurer: claim submission confirmations, denial letters, phone call dates and representative names, and all appeal correspondence. Under federal law, you have the right to know the specific reason your claim was denied and the exact medical criteria applied. If you suspect an AI made the decision, explicitly ask in writing whether a licensed physician at the insurer reviewed your case before the denial was issued. You can also file a formal complaint with your state insurance commissioner — regulators like Washington's Mike Kreidler have been vocal about the need for accountability in AI-driven risk assessment. Documentation is your most powerful tool.
Frequently Asked Questions
Can a health insurance company legally deny my claim using an AI algorithm without a licensed doctor reviewing it?
This is the precise question being tested in federal courts right now. Insurers are generally required under state and federal law to involve licensed clinicians in medical necessity determinations, but the rules around automated systems are still evolving. The ongoing class action lawsuits against Cigna and UnitedHealthcare are directly challenging whether AI-only denials violate insurers' legal obligations to their members. Until clearer federal regulation exists, you have the right to request written confirmation that a human physician reviewed your case before any denial of your policy coverage was issued — and to appeal immediately if one did not.
How do I appeal a health insurance claim denial made by an AI system in 2026?
Start by obtaining the denial letter in writing, which must specify the clinical criteria used and the reason for denial. File a formal internal appeal with your insurer within the deadline in your plan documents. If that is denied, request an external independent review — a right guaranteed under the Affordable Care Act. For Medicare Advantage denials specifically, contact your State Health Insurance Assistance Program (SHIP) for free, unbiased counseling. Document every step and communicate in writing whenever possible. Given that 80% of Cigna Medicare Advantage AI-driven denials that were appealed were ultimately overturned, the effort is almost always worth making.
Does AI-driven claims denial affect Medicare Advantage policy coverage differently than traditional Medicare?
Yes, and the difference is significant. Medicare Advantage plans are administered by private insurers — UnitedHealthcare, Humana, Cigna, and others — and it is these private plans that have been most aggressive in deploying AI for utilization management and claims management. The 2024 Senate Subcommittee report specifically called out Medicare Advantage insurers for using algorithmic tools to deny care at rates far exceeding what human reviewers would produce. Traditional Medicare, run directly by the federal government, does not use these same AI denial systems. If you are doing an insurance comparison between Medicare Advantage and traditional Medicare, how each handles prior authorization and claims is a critically important factor to weigh.
What is the nH Predict algorithm and how could it affect my health insurance claims?
nH Predict is an AI tool developed by NaviHealth, a subsidiary of UnitedHealthcare, designed to predict how long a patient should need post-acute care such as skilled nursing or home health services after a hospital stay. A federal class action lawsuit alleges the tool had a known 90% error rate and was used to override treating physicians' recommendations for elderly Medicare Advantage patients. UnitedHealthcare's skilled nursing facility denial rate jumped from 1.4% in 2019 to 12.6% in 2022, coinciding directly with nH Predict's rollout — a pattern that drew scrutiny both in the 2024 Senate report and in ongoing litigation. Researchers have characterized this kind of automated risk assessment as a fundamental shift in who actually decides what care you receive.
Can switching health insurance plans help me avoid AI claim denials and get better insurance savings in 2026?
Switching plans can reduce your exposure to aggressive AI-driven denials, but there is no guarantee. Different insurers use different AI tools and have meaningfully different historical denial rates. When doing an insurance comparison, look for plans with lower prior authorization denial rates (available from CMS for Medicare Advantage), higher appeal overturn rates, and fewer consumer complaints filed with your state insurance regulator. Keep in mind that a plan promising significant insurance savings on the premium may offset those savings through a more aggressive claims management approach that denies a higher share of your care. Always consult a licensed insurance agent before changing plans, especially if you have ongoing medical needs that require frequent prior authorization.
Disclaimer: This article is for informational purposes only and does not constitute insurance advice. Always consult a licensed insurance agent for personalized guidance.
No comments:
Post a Comment