The administrative landscape of healthcare is undergoing a profound transformation, as artificial intelligence (AI) increasingly infiltrates the complex and often contentious process of prior authorization. What was once a manual, human-driven endeavor is rapidly evolving into an automated exchange between sophisticated algorithms, leaving physicians and patients caught in the crossfire of "bot-to-bot" communication. This shift, while promising potential efficiencies, is also generating significant frustration and exacerbating existing challenges within the healthcare system.
The core of the issue lies in how AI, particularly large language models, are being employed by health insurance companies to review medical necessity and approve or deny claims. While proponents argue AI can streamline processes and reduce administrative burdens, the reality for many healthcare providers is a surge in denials, often perceived as arbitrary and based on a superficial understanding of patient documentation.
One stark illustration of this phenomenon comes from Dr. Yolanda Troublefield, an attending surgeon and otolaryngologist at Southcoast Physicians Group in Massachusetts. A member of the American Academy of Otolaryngology–Head and Neck Surgery’s Physician Payment and Policy Workgroup, Dr. Troublefield recently encountered a perplexing denial for an eight-year-old child requiring an adenoidectomy with ear tube placement. The child had a history of recurrent otitis media, with clear documentation from both primary care physicians and otolaryngologists, supported by an audiogram. Even the patient’s parents, both physicians themselves, understood the standard of care.
"We’ve been doing tonsils and adenoids and ventilation tubes for the last 50 years—seems pretty straightforward," Dr. Troublefield recounted. "Every single ‘i’ has been dotted, ‘t’ has been crossed." Yet, the prior authorization request was denied. Dr. Troublefield suspects the involvement of AI is to blame, citing the algorithm’s apparent inability to interpret the entire clinical note holistically. In this case, while the diagnosis was "recurrent otitis media," the assessment-and-plan section of the note included "acute otitis media," intended to signify the immediate need for antibiotics during that specific visit. The AI, however, seemingly fixated on this phrase, leading to the denial and necessitating a time-consuming peer-to-peer conversation, a process that disrupts her schedule and delays patient care.
Denials at Machine Speed: The Rise of AI-Driven Rejections
The integration of AI into prior authorization processes, and even for clawbacks on services already rendered, has led to an unprecedented acceleration of claim denials. Otolaryngologists report receiving rejections with alarming speed and often in ways that defy logical medical reasoning. This forces physicians to invest more time and resources in securing coverage for necessary patient care, leading to mounting exasperation. Health practices are increasingly turning to AI themselves to manage documentation submissions and respond to denials, creating automated cycles with potentially far-reaching and as-yet-unknown consequences for the healthcare ecosystem.
"What we’re seeing is faster denial of claims because of the use of AI," Dr. Troublefield observed. "You can set the dial in different ways. You can set it high, medium, or low. And they’re set high for denials." This sentiment is echoed in broader physician surveys. A report by the American Medical Association (AMA) last year revealed significant concerns among physicians regarding AI’s impact on prior authorization. Sixty percent of physicians expressed worry that AI has already increased or will increase prior authorization denial rates. Furthermore, 55% stated that prior authorization delays access to necessary care either always or often, and a striking 79% reported that prior authorization sometimes leads to patients abandoning treatment altogether. Alarmingly, 26% of physicians indicated that prior authorization has resulted in a serious adverse event for a patient under their care.
Dr. Bruce Scott, a former president of the AMA and an otolaryngologist at ENT Care Centers in Louisville, Kentucky, shared his initial optimism about AI’s potential to alleviate administrative burdens. "I think physicians were very hopeful that AI was going to be a boost, a solution if you will, to reduce the administrative burdens that we all face," he said. However, this hope has largely been unmet. Dr. Scott foresees a future where automated systems clash directly: "Eventually, their bot is going to talk to my bot. Because my bot is going to tell me how to document so I get authorization, and then their AI is going to get even smarter and deny that, so I get a denial letter back. And in the meantime, the physicians and patients are stuck in the middle, and that’s the problem."
Regulators Enter the AI Fight: Charting a New Course
In response to the escalating issues, a wave of new state laws is being enacted or is slated to take effect, aiming to regulate the use of AI in claim denials. Alabama has introduced legislation requiring insurers to base determinations on a patient’s unique circumstances rather than relying solely on group datasets. Indiana has prohibited insurers from using AI as the sole basis for downcoding a claim. Similarly, a Washington state law mandates that claim determinations can only be made by licensed and qualified health professionals. These legislative efforts represent a growing recognition of the need for human oversight and patient-centric decision-making in the face of automated processes.
Beyond state-level interventions, the federal government is also preparing to deploy AI within its own health insurance programs, specifically through Medicare. The WISeR (Wasteful and Inappropriate Service Reduction) program is designed to leverage AI, in conjunction with human review, to "reduce clinically unsupported care by working with companies experienced in using enhanced technologies to expedite and improve the review process for a pre-selected set of services that are vulnerable to fraud, waste, and abuse," according to the Centers for Medicare and Medicaid Services (CMS). This program, launched in January in six states—New Jersey, Ohio, Oklahoma, Texas, Arizona, and Washington—signifies a federal acknowledgment of AI’s potential, albeit with a focus on mitigating waste and abuse.
Fighting Automation with Automation: Physicians Adapt to the AI Landscape
The pervasive use of AI by insurers has compelled physicians to adopt similar technologies to navigate the evolving system. Dr. Bradford Bichey, a rhinologist at Indiana Sinus Centers, has developed an AI-powered product designed to streamline physician office operations, including the critical task of generating documentation for prior authorization submissions. Dr. Bichey observed a significant increase in "clawback attempts" a few years ago, where insurers retroactively questioned coverage for previously approved services. He went from receiving approximately five such requests periodically to encountering dozens at a time. The language used in these requests, he noted, often suggested "early experimentation" with AI, appearing "slightly off" and, in one instance, "almost hateful." While the tone has since become more professional, the underlying challenges persist.
In response, Dr. Bichey created "Blue," an AI product developed under his company, Nemedic. This software not only generates clinical notes based on patient encounters but also crafts prior authorization documents precisely tailored to the requirements of specific insurers. For example, if "United" is mentioned during a patient visit, the software automatically cues the generation of a document formatted to meet United Healthcare’s stipulations, drawing from the insurer’s publicly available guidelines. These submissions often require extensive justification beyond the scope of a typical clinical note. For a nasal endoscopy, the AI might generate a detailed explanation such as, "routine anterior examination was insufficient to visualize the deeper nasal passage and assess for ongoing mucosal disease. Diagnostic rigid nasal endoscopy was medically necessary to direct visualization of the sinonasal mucosa," while also carefully noting exclusions like, "this procedure was not performed for routine screening, but to inform management."
"You have to know [the insurer’s] inclusion and exclusion," Dr. Bichey emphasized. "And it’s almost impossible for someone who does billing to constantly keep up with all these plans because we see hundreds of different types of insurance." His AI solution simplifies the appeal process, which often involves highlighting overlooked information in the original submission. "When we do this, we just resubmit the same note and say, ‘You’re wrong—look at paragraph three,’" Dr. Bichey explained. He reports not having to engage in a peer-to-peer call for two years due to the significantly enhanced upfront documentation facilitated by his AI tool.
Not an AI Problem, but a Governance Problem: The Underlying Tensions
Despite the apparent role of AI in claim denials, some experts argue that the technology itself is not the root cause of the problem. Dr. Matthew Crowson, assistant professor of otolaryngology–head and neck surgery at Harvard Medical School and director of clinical informatics and artificial intelligence at Massachusetts Eye and Ear, contends that the issue lies more with the governance, or lack thereof, surrounding AI implementation.
"It’s not so much that AI is the problem," Dr. Crowson stated. "It’s more the governance around this stuff, or lack thereof." He posits that AI has not resolved the inherent tension between providers and payers. "It’s turning into a battle of the bots," he observed. "The application of AI by providers and payers is cutting both ways. It doesn’t resolve the fundamental tension between approving and denying based on medical necessity. It’s scaling up existing decision-making workflows faster."
However, Dr. Crowson also acknowledges AI’s potential to improve administrative efficiency in ways that benefit patients. For instance, when a prescribed medication is not on a health insurance plan’s formulary, a nurse or medical assistant would typically spend considerable time gathering supporting information and completing appeal forms. AI can now draft a contextualized appeal rapidly, potentially expediting medication delivery to patients. "What would take a human process maybe seven business days to do, now you can do it in one," he noted. "I think administratively, it’s massively speeding up our ability to respond quickly."
Efficiency Gains—or Just More Activity? The Unclear Impact of AI
A report by the Peterson Health Technology Institute earlier this year, based on workshops involving a diverse group of healthcare leaders, presented a nuanced and somewhat unclear picture of AI’s impact on the U.S. medical landscape. While AI may reduce the cost of prior authorizations for individual organizations, it has not demonstrably lowered overall system costs. The report also indicates that provider adoption of AI is contributing to increased "billing intensity," with more severe diagnoses and advanced treatments being identified, thereby driving up associated costs.
Workshop participants expressed concerns that the optimized use of AI by both insurers and providers could lead to a scenario where the entire process becomes "more activity-intensive" rather than truly more efficient. One healthcare provider, quoted in the report, lamented, "Bots don’t get tired of asking questions, so my review queue keeps growing." This suggests that the sheer volume of automated interactions, even if efficient for the bots, may overwhelm human capacity and fail to achieve genuine systemic efficiency.
Patients Caught in the Middle: The Human Cost of Automated Decisions
Meanwhile, the prior authorization process, amplified by AI, continues to strain the physician-patient relationship. Dr. Troublefield highlights the daily challenge of engaging in complex conversations with patients about insurance coverage hurdles. "It takes more time—I don’t think any of us runs on schedule at all," she lamented. "I would say five years ago, we probably could run on schedule. Not anymore. It never, ever, ever happens." She added, "We’re finding that we’re constantly having to apologize, apologize, apologize. And that’s not the way people want to be treated."
Dr. Scott recounted a distressing instance where a patient with a maxillary sinus tumor, who had finally agreed to surgery, received a denial letter from her insurer stating the procedure was denied because she hadn’t yet been on an antibiotic. This assessment was, in Dr. Scott’s view, "preposterous." The patient, influenced by the insurer’s communication, then questioned the necessity of surgery, asking, "The insurance company said I might get better on an antibiotic, so shouldn’t we try an antibiotic?" Dr. Scott expressed the difficulty of such situations: "Now I’ve got to go explain to this patient, and I’m thinking about the trust relationship that I had with this patient and having to re-establish that trust."
He advocates for physicians to appeal denials more frequently, noting that his practice appeals every single one. However, he acknowledges the pervasive issue of physician burnout, which can leave practitioners feeling unable to engage in constant battles. Dr. Scott and the AMA are actively lobbying for physician involvement in the development of regulations and processes surrounding AI in healthcare, emphasizing the need to address practical concerns like practice viability, potential litigation, and reimbursement.
A Fork in the Road for Prior Authorization: Navigating the Future
Dr. Crowson views the current situation as a critical juncture for prior authorization. He outlines two potential paths: one where both providers and payers utilize AI to process routine paperwork and basic tasks more rapidly, freeing up human resources for complex cases. The alternative is an "escalating compliance and gaming war," where the focus shifts to outmaneuvering automated systems.
He expresses hope for a "pragmatic, middle-of-the-road approach," advocating for the use of AI for its strengths: automating administrative monotony and simple tasks that can be safely and efficiently scaled. Dr. Troublefield, while concerned that AI might introduce new requirements for physicians to use specific "buzzwords," finds the concept of software that automatically generates insurer-compliant templates attractive, provided it is practical.
She remains cautiously optimistic that technological advancements, coupled with "smart legislation" that regulates AI’s role in insurance decisions, could offer solutions. The inclusion of more patient voices, and even celebrity endorsements, could significantly bolster legislative efforts. Ultimately, however, Dr. Troublefield believes physicians must adapt to the changing landscape. "The system is changing, and either you have to be part of the change, or you’re going to get left behind," she concluded. "It’s the same thing as when the automobile happened, right? You’re not stopping it, so you have to figure out ways within the system to make sure that what you want, and what your goals are, are actually achieved." The "battle of the bots" is just beginning, and its outcome will profoundly shape the future of healthcare access and delivery.
