How AI Can Help You Appeal Health Insurance Denials

Tom Whitehead knows which medicine works to treat his Crohn’s disease. But when he switched insurers in 2025, he says was denied coverage for the expensive medication, Remicade, which he received in a monthly infusion. He was instead put on what’s called a biosimilar, which is essentially a slightly different version of the drug made from the same materials.

The biosimilar did not work well for him. His symptoms worsened, he had urgent and unpredictable bowel movements, and he got kidney stones twice, ending up in the hospital, he says. His doctor’s office appealed the coverage denial after his bad side effects, he says, but the company upheld it and he could not get the medication he wanted.

Whitehead found Warris Bokhari, the founder of Claimable, an AI tool that helps people appeal insurance denials for medication and other treatments. Claimable drafted a letter for him and told him to send it to. With his input, Claimable outlined his experience with the biosimilar drug, detailed how he’d missed work because of his symptoms, and cited federal statutes that, the AI argued, meant that his insurer had to cover this “medically necessary drug.”

Health insurers are increasingly using AI to evaluate requests for coverage of drugs and medical procedures. But patients, too, are using AI to fight back against denials. Services like Claimable, Fight Health Insurance, Mediloop, and Billyze help people understand their medical bills and push back against what could be improper billing or wrongful denials.

Many Americans have had an experience where their insurance denies a procedure or medication request from a doctor. But very few of them appeal those denials. One recent survey of more than 4,500 U.S. working-age adults with private health insurance found that 21% reported that they or a family member had experienced a denial of coverage in the past year. Only about half of them appealed it.

People often don’t appeal insurance denials for several reasons; it can be intimidating to know what paperwork to file, and the bills can start to pile up, says Sara R. Collins, a senior scholar at the Commonwealth Fund and one of the authors of the report. In the survey, some people who appealed their denial said that their provider had threatened to send their overdue bills to collections while they awaited word from their insurer.

About 43% of people in the study said that a claim denial led them to incur medical debt, and half of those bills were $1,000 or more. “It’s not insignificant sums of money when you end up with a denial of care,” she says.

Often, patients feel helpless when they receive a medical bill or prior authorization denial, says Collins. They don’t know who to contact or what to do. Though their provider can sometimes help with next steps, that isn’t always the case. AI tools have emerged to offer guidance, usually for a fee.

Each of the AI tools available to patients does something slightly different. Claimable helps people appeal insurers’ decisions to deny coverage for drugs or medical care and costs about $40 per appeal. Fight Health Insurance is a free tool that helps patients write letters appealing denials. Billyze helps search hospital bills for possible errors and helps patients question charges that may be erroneous. It costs $79 per report, though for bills over $1,000, it promises to refund the service fee if its analysis finds no billing discrepancies. Mediloop also reviews bills for potential errors and negotiates with providers to reduce what patients owe. It costs $129 to negotiate a bill and also offers an annual household plan.

“I realized that if you could invert and and actually give patients the ability to argue articulately against the insurer, you might actually change someone’s life,” says Bokhari, the CEO and co-founder of Claimable.

Using AI to help with medical bills only became realistic in the last few years, says Holden Karau, one of the founders of Fight Health Insurance. Open generative models—AI systems with publicly available code that users modify to their own purposes—only became good enough to help build this kind of system in early 2024, Karau says.

“That’s when the models got good enough that it was like, yeah, we can actually do this, let’s make it available for people to use,” she says.

It can be difficult to build an AI model that helps people appeal denials because the insurer—not the patient—usually has most of the information about why they’re denying claims and what makes them overturn their appeals, say Karau and her co-founder Melanie Warrick. But the two still found data to build into their model, including published information from state regulators about what language they used to overturn insurers’ denials.

For instance, if an insurance company said that a patient needs to get an X-ray before they get an MRI, and a regulator in California overturned that decision, the AI model could use the language the state regulator used to overturn that denial in its appeal letter.

Like Karau, Flavia Bojescu, the founder of Mediloop, became interested in pushing back on health insurance bills after getting into an accident in the U.S. and stressing over how much her medical bills would be, even though she had insurance at the time. Bojescu says a physician in the emergency room asked her if she wanted stitches, which she didn’t understand. In her home country of Romania, the doctor would decide whether or not she needed stitches; cost or coverage wouldn't be factors for the patient to consider.

Bokhari says he hopes that someday, his AI tool won’t need to exist. If it starts getting more expensive for insurers to deny care than to just pay for it, maybe they’ll approve more claims, he says.