In recent discussions surrounding healthcare, particularly Medicare, concerns have surfaced regarding the introduction of artificial intelligence (AI) in claims processing. The pilot program initiated by the Trump administration aims to streamline decision-making but has raised alarms about potential implications for patient care. Critics argue that this new initiative could equate to an increase in denied claims, ultimately compromising the quality of care for Medicare beneficiaries.
Yves here. This troubling article highlights how the deployment of AI in the processing of Medicare claims seems designed to increase claim denials. Although the program is set to begin as a pilot, it remains unclear how it will function. However, it appears that it will often necessitate prior approval, a practice that can hinder or delay treatment, especially for significant procedures such as surgeries.
Another objective may be to encourage retirees to shift from traditional Medicare to Medicare Advantage. As previously noted, Medicare Advantage operates as a second-tier system, frequently offering lower or no premiums at the cost of reduced coverage, often relying on restricted provider networks.
Yes, they really do want to kill us.
By Lauren Sausser and Darius Tahir. Originally published at KFF Health News
The Trump administration plans to roll out a new program aimed at determining how much money an AI algorithm can save the federal government by denying treatment to Medicare patients. The pilot initiative, which aims to eliminate so-called wasteful or “low-value” services, signifies a federal expansion of the contentious prior authorization process. This requires either patients or their healthcare providers to obtain insurance approval before going ahead with specific procedures, tests, and prescriptions. The program will impact Medicare patients, along with the medical providers serving them, in several states including Arizona, Ohio, Oklahoma, New Jersey, Texas, and Washington, commencing on January 1 and extending through 2031.
This development has caught the attention of both politicians and policy experts. Traditional Medicare, which serves adults aged 65 and older and those with disabilities, has generally avoided prior authorization. However, this practice has become common among private insurance providers, particularly within the Medicare Advantage sector.
The timing of the announcement raised eyebrows: it was revealed just days after the Trump administration promoted a voluntary initiative for private health insurers aimed at minimizing their use of prior authorization, which often leads to significant delays in patient care, as noted by Mehmet Oz, the head of the Centers for Medicare & Medicaid Services.
“It erodes public trust in the health care system,” Oz remarked during a press conference. “We cannot tolerate this in this administration.”
Yet critics, such as Vinay Rathi, a physician and policy researcher from Ohio State University, argue that the Trump administration’s dual approach sends mixed messages. “On the one hand, the federal government seeks to adopt cost-saving measures from private insurers,” he stated. “On the other hand, it reprimands them for their methods.”
Concerns have been voiced by patients, healthcare providers, and lawmakers alike regarding what many see as “delay-or-deny tactics,” which can obstruct or postpone access to necessary care, thereby causing irreparable harm and even fatalities.
“Insurance companies have adopted a business model where they collect premiums from patients only to minimize their payout to healthcare providers,” said Rep. Greg Murphy, a Republican from North Carolina and a urologist. “This is a common practice in every insurance company.”
While insurers claim that prior authorization is a method to curb fraud, waste, and potential harm, public dissatisfaction with insurance denials has surged. In December, the notable shooting death of the CEO of UnitedHealthcare resulted in significant public outcry, with many considering the alleged perpetrator a folk hero.
Data from a July poll by KFF, a health information nonprofit group, indicated widespread public discontent regarding prior authorization, with nearly three-quarters of respondents identifying it as a “major” issue.
Oz mentioned in a June press conference that public unrest was partly responsible for the administration’s focus on prior authorization reform within the private insurance sphere.
Nonetheless, the administration is proceeding with the expansion of prior authorization in Medicare. Alexx Pons, a spokesperson for the Centers for Medicare & Medicaid Services (CMS), pointed out that both initiatives aim to “protect patients and Medicare funds.”
Unanswered Questions
The pilot program, known as WISeR (Wasteful and Inappropriate Service Reduction), will evaluate AI algorithms in determining prior authorization outcomes for certain Medicare services, including skin and tissue substitutes, electrical nerve stimulators, and knee arthroscopy.
The government claims these procedures are particularly susceptible to “fraud, waste, and abuse,” making them suitable for prior authorization. Additional services may eventually be included in this program, but inpatient-only, emergency, or procedures presenting significant risks if delayed will be exempt from AI evaluations, according to the official announcement.
Although the concept of using AI in healthcare is not groundbreaking, Medicare has historically been hesitant to adopt these techniques, unlike private insurance sectors. Past Medicare practices involving prior authorization have utilized contractors not incentivized to deny services. Experts express concern that the federal pilot could change this dynamic.
Pons assured KFF Health News that no Medicare requests would be rejected without prior review by a “qualified human clinician,” while vendors would be barred from any payment arrangements linked to denial rates. Though vendors may be compensated for savings, Pons stated that numerous safeguards would ensure “no incentives to deny medically appropriate care.”
However, as noted by Jennifer Brackeen, senior director of government affairs for the Washington State Hospital Association, the shared savings models could create a strong motivation for vendors to reject necessary care.
Doctors and policy experts have voiced additional concerns. Rathi indicated that the initiative “is still underdeveloped” and relies on “messy and subjective” assessments. He fears that the model’s success hinges on contractors evaluating their own performance, a decision fraught with potential bias.
“I’m unclear about how they plan to determine whether this initiative is benefiting or harming patients,” he stated.
Pons asserted that the AI use in the Medicare pilot would receive “strict oversight to guarantee transparency, accountability, and adherence to Medicare regulations and patient safeguards.” He emphasized that CMS is committed to ensuring automated tools complement clinically sound decision-making rather than replace it.
Experts unanimously agree that AI has the potential to streamline a currently cumbersome process that often leads to harmful delays and denials in patient care. Health insurers maintain that AI can mitigate human error and bias while being more cost-effective for the healthcare system. They also argue that humans ultimately make the final coverage decisions, not algorithms.
Still, some experts question how frequently this actually occurs. “There appears to be a lack of clarity about what constitutes ‘meaningful human review,’” observed Amy Killelea, an assistant research professor at Georgetown University’s Center on Health Insurance Reforms.
A report published by ProPublica in 2023 revealed that doctors at Cigna, who assessed payment requests, averaged only 1.2 seconds per case during a two-month review period.
Cigna representative Justine Sessions clarified that the company does not deploy AI to deny care or claims. Instead, the ProPublica report described a “simple software-driven process” aiding prompt payments for commonly utilized, lower-cost treatments that do not involve AI.
Nevertheless, class-action lawsuits against major health insurers have alleged that deficient AI models disregard doctor recommendations and fail to recognize individual patient needs, imposing a financial burden on patients seeking care.
Additionally, a survey released by the American Medical Association in February found that 61% of physicians believe AI is “exacerbating prior authorization denials, causing avoidable patient distress and escalating unnecessary waste.”
Chris Bond, a spokesperson for the insurers’ trade group AHIP, asserted that the organization is “fully focused” on executing the commitments made to the government, which include minimizing the scope of prior authorization and enhancing clarity in patient communications regarding denials and appeals.
‘This Is a Pilot’
The Medicare pilot program highlights persistent anxieties about prior authorization while introducing new concerns. While private insurers have been opaque regarding their AI implementations and prior authorization practices, researchers are inclined to believe that these algorithms are often programmed to automatically deny high-cost services.
“The higher the price tag, the greater the likelihood it will be denied,” remarked Jennifer Oliva, a professor at the Maurer School of Law at Indiana University-Bloomington, who studies AI regulation and health coverage.
Oliva elaborated in a recent paper for the Indiana Law Journal that when a patient’s life expectancy is in question, insurers are motivated to rely on AI assessments. As time passes and patients or their healthcare providers appeal denials, the risk of the patient’s death during this process increases. The longer the appeal, the less likely insurers are to approve the claim, she noted.
“The primary goal is to make access to high-cost services extremely challenging,” she indicated.
As the implementation of AI by health insurers continues to rise, these algorithms present a “regulatory blind spot” that necessitates increased scrutiny, according to Carmel Shachar, a faculty director at Harvard Law School’s Center for Health Law and Policy Innovation.
The WISeR pilot represents “an interesting development” in utilizing AI for ensuring that Medicare funds are allocated to high-quality healthcare, she noted. However, uncertainties surrounding the details complicate the evaluation of its effectiveness.
Lawmakers are wrestling with similar questions. “What methodology is being employed to evaluate this initiative? How will you ensure it won’t deny necessary care or cause higher rates of care denial?” asked DelBene, who has signed an August letter to Oz seeking clarification regarding the AI initiative. Bipartisan concern exists; Murphy, who co-chairs the House GOP Doctors Caucus, acknowledged that numerous physicians fear the WISeR pilot could encroach upon their medical practices if the AI denies care recommended by doctors.
In a recent development, House members from both parties backed a proposal put forth by Rep. Lois Frankel, a Democrat from Florida, aimed at blocking funding for the pilot in the fiscal 2026 budget for the Department of Health and Human Services.
While AI is undeniably becoming a fundamental element of healthcare, Murphy remarked that the effectiveness of the WISeR pilot in saving Medicare funds or exacerbating existing issues associated with prior authorization remains to be seen. “This is a pilot, and I’m open to observing its outcomes,” he said. “However, I will always prioritize the perspective that physicians know what’s best for their patients.”
