AI is denying medical claims

UnitedHealth, Humana, and Cigna are using AI programs to reject insurance claims that doctors say patients need, sparking class-action lawsuits from patients who lost coverage.

The real threat from AI

While headlines focus on AI extinction risks, the immediate danger is happening now: algorithms are deciding whether people get medical care, often without their knowledge.

“This is AI's threat to people's health right now.”

How AI claims review differs

Old insurance algorithms made rough estimates based on age and health factors. New AI systems make specific predictions about future health to approve or deny individual claims, often opaquely.

The human review problem

Federal law requires a human doctor to make final claim decisions, but it's unclear whether insurers are using AI to support doctors or replace them entirely.

“The difficulty is distinguishing between AI supporting meaningful human processing and AI basically replacing the human with the human just passively defaulting to the AI judgment.”

Gene Lokken's story

After a fall, the 91-year-old improved with nursing care. But UnitedHealth's AI algorithm denied coverage after two weeks, forcing his family to pay $12,000 to $14,000 monthly. He died in the facility a year later.

The nH Predict pattern

UnitedHealth's AI tool, called nH Predict, typically starts denying nursing home claims after two weeks, regardless of whether patients are still recovering.

More patients harmed

Glennette Kell couldn't walk but lost coverage. Darlene Buckner had an infected heart wound but was denied hospital care. Both had to pay thousands out of pocket.

UnitedHealth's defense

The company claims federal law requires human doctors to make final decisions and says employees are instructed to follow the law, but plaintiffs allege workers face discipline for deviating from AI predictions.

Humana's similar lawsuit

Humana faces its own lawsuit over the same nH Predict algorithm. Patient Joanne Barrows, 86, had claims denied after just two weeks of rehab despite doctor's orders not to bear weight.

Cigna's batch denials

Cigna used a different AI algorithm that allowed doctors to deny hundreds or thousands of claims at once if they didn't match preset criteria, bypassing individual review.

Patients unaware of AI

Most patients had no idea AI was making decisions about their coverage until they were told to leave their care facility or received a denial notice.

“They just had no idea what was going on or why this was being done to them.”

Medicare joins the trend

In June 2025, the Trump administration announced WISeR, an AI program to make prior authorization decisions for traditional Medicare in six states, affecting millions of beneficiaries.

WISeR's troubled rollout

Since launching in January, the WISeR model has caused approval delays and denials for covered services. One review took eighty-three days instead of the required seventy-two hours.

Doctors report patient suffering

Physicians documented patients calling in tears because pain procedures were delayed. One Arizona patient couldn't exercise due to denied steroid injections for chronic back pain.

“We have patients calling our offices crying in pain because their procedures are being delayed while awaiting approvals or guidance tied to this model.”

Aggressive denials revealed

Freedom of Information Act documents showed one company using WISeR denied more prior authorization requests than it approved, contradicting the program's stated purpose.

The regulation gap

AI is being deployed faster than regulators can oversee it. There's no clear framework for what keeping a human meaningfully in the loop actually means in practice.

“I don't think that there's anything I've seen that has effectively defined what it means for a human to be meaningfully in the loop.”

Some policy efforts underway

Several states have adopted transparency rules for AI in insurance, and Congress proposed the bipartisan Protecting Patients from Automated Denials Act, but enforcement remains unclear.

What you can do now

Appeal denied claims. Most get overturned on appeal. Watch out for vulnerable people in your life. Push elected officials to regulate AI use in healthcare.

“You do not necessarily have to accept that the first answer from your insurance company is the final determination.”