A new study reveals that hospitals used artificial intelligence tools to drive up healthcare spending by nearly a billion dollars — and American families are footing the bill through higher premiums and out-of-pocket costs.
The findings add fuel to a growing national debate over how AI technology is reshaping the economics of American healthcare, raising questions about whether automation is being used to improve patient outcomes or simply to maximize revenue. As hospitals and insurance companies race to deploy sophisticated AI systems, the financial consequences are landing squarely on consumers who have little visibility into these behind-the-scenes billing practices.
Analysis from the Blue Cross Blue Shield Association (BCBSA) shows that a major rise in patients being documented as having complex conditions led to $942 million in additional medical costs for Blue Cross and Blue Shield companies between 2023 and 2025.
The catch? Treatment levels didn’t actually increase.
This discrepancy between documented diagnoses and actual medical interventions lies at the heart of concerns about AI-enabled medical coding systems. These tools scan patient records and identify conditions that can justify higher reimbursement rates under Medicare and private insurance payment models, which use diagnosis codes to determine how much providers are paid. When secondary conditions are added to a patient’s chart, even without any change in treatment, the hospital can move that claim into a more lucrative payment category.
“If patients are truly sicker, we’d expect to see more treatment. For example, we’re seeing significantly more anemia diagnoses at these hospitals without a corresponding increase in transfusions. The disconnect between diagnoses and treatment suggests that AI is identifying more billable conditions, not sicker patients.”
BCBSA Senior Vice President of Product and Data Science Luke Chalker made that statement, according to the association’s report, which raises serious concerns about AI-fueled billing practices inflating U.S. healthcare spending.
The anemia example Chalker cited illustrates a pattern that appears across multiple diagnostic categories. When AI systems identify conditions like malnutrition, kidney disease, or diabetes complications in patient records, those diagnoses increase the complexity score assigned to that hospital stay. Under Medicare’s payment system and many commercial insurance contracts, higher complexity scores translate directly into higher payments, creating a financial incentive to document every possible condition whether or not it affects treatment decisions.
The research shows roughly $650 million of the increased costs can be traced to secondary diagnoses that shifted claims into higher-reimbursement categories — billing tricks that have nothing to do with actual patient care.
That $650 million figure represents the portion of costs driven purely by coding changes rather than any measurable increase in patient severity or treatment intensity. The remaining portion of the $942 million reflects other AI-influenced billing adjustments. These secondary diagnoses often appear in patient records without triggering any additional tests, medications, or interventions, suggesting their primary purpose is financial rather than clinical.
Hospital AI found more diagnoses to bill. Insurer AI learned to deny them. Nearly $1B moved, treatment didn’t rise, and patients fund both systems through higher premiums. The machines got smarter. The patient just got a bigger bill. pic.twitter.com/3Uq0jpwLSS
— Migo (@ReiteConMig0) September 27, 2026
Insurers and consumer advocates warn that these AI-driven billing games will ultimately push insurance premiums and out-of-pocket expenses even higher for families and employers already struggling with skyrocketing healthcare costs.
The concern extends beyond just the immediate dollar figures identified in the BCBSA analysis. When hospitals systematically increase their reimbursement rates through AI-enabled coding, insurance companies must raise premiums across their entire customer base to cover those costs. Employers offering health benefits face higher contributions, often leading them to shift more costs onto workers through higher deductibles and copayments. Self-employed individuals and families buying insurance on their own face the steepest increases, with no employer to share the burden.
BCBSA Senior Vice President of External Affairs David Merritt said, “We know families are frustrated by the rising cost of healthcare, and we are committed to tackling the root causes of these higher prices.”
Meanwhile, hospitals and groups like the American Hospital Association are firing back — accusing insurance companies of deploying their own AI systems to automatically deny or downcode legitimate claims without human review.
This counterargument reflects the increasingly adversarial relationship between hospitals and insurers, with both sides investing heavily in AI technology to gain an edge in billing disputes. Hospital representatives argue that AI coding tools simply ensure that all legitimate diagnoses are properly documented, something that manual coding processes often missed. They contend that insurers are using the BCBSA findings to justify their own aggressive denial practices, creating a technological arms race where patients are caught in the middle.
The fight over AI in healthcare could also shake up November’s midterm elections. A recent NBC News survey found that U.S. voters trust Democrats over Republicans on healthcare by 22 percentage points — a political vulnerability Republicans can’t afford to ignore as families watch their medical bills climb.
Healthcare costs consistently rank among voters’ top concerns, and stories about AI systems being used to inflate hospital bills could reinforce public perception that the healthcare system prioritizes profits over patients. As both parties craft their midterm messaging, the question of who will rein in AI-driven billing practices may become a defining issue for swing voters facing their own rising premiums and medical debt.
