A major new study from the Blue Cross Blue Shield Association (BCBSA) has put a dollar figure on a concern that has quietly grown inside the insurance industry. Hospital AI billing tools may be adding billions to the nation’s healthcare tab — and the data now backs it up.
What the BCBS Study Found
BCBSA’s data analytics arm, Blue Health Intelligence (BHI), conducted the research. Analysts reviewed commercial inpatient claims from tens of thousands of maternity admissions nationwide. The study covered a three-year window ending in March 2025.
A Focus on One Key Diagnosis
Researchers zeroed in on acute posthemorrhagic anemia — a condition that typically requires clinical intervention such as blood transfusions. They found that diagnoses of this condition surged dramatically at a specific subset of hospitals. However, the spike in coding did not come with a matching rise in actual treatment.
In other words, hospitals billed for the condition far more often. Yet patients were not receiving the care that typically accompanies it. This gap between coding and treatment is the heart of the problem.
The Upcoding Problem Explained
Upcoding is the practice of billing for more complex or severe conditions than those actually treated. It results in higher reimbursements for hospitals. Moreover, it pushes costs onto insurers, employers, and ultimately patients.
How AI Tools Enable Upcoding
Many hospitals now use AI-powered clinical documentation tools. These systems rely on ambient listening technology. They also analyze lab reports and physician notes to automatically assign billing codes.
BCBSA reviewed hospitals where either the facility or a vendor had publicly disclosed AI use. At one such hospital, overall case complexity jumped 6.7% following that disclosure. Meanwhile, other hospitals in the same state saw complexity grow by less than 1% over the same period.
Furthermore, the top 10% of hospitals in the study drove the bulk of the increases. By March 2025, nearly 60% of inpatient admissions that could be coded as complex were handled at those facilities — up from about 47% in April 2022.
How Much Is AI Billing Really Costing Americans?
The financial impact is striking. First, consider the maternity care example. The increased coding of acute posthemorrhagic anemia alone added $22 million to maternity-related hospital costs in a single year across the facilities studied.
The National Picture
Projected nationally, the numbers are far larger:
- Inpatient excess spending: approximately $663 million
- Outpatient exposure: at least $1.67 billion
- Total estimated impact: approximately $2.3 billion
Additionally, within one participating BCBS plan’s commercial inpatient population, per-member costs rose 9% between 2023 and 2024. Coding intensity alone contributed roughly 20% of that increase.
Luke Chalker, chief product officer at BHI, put it plainly. “When coding changes and it drives a reimbursement change, both a payer is paying more and a member is paying more — from an affordability perspective, through out-of-pocket maximums and through their deductibles,” he told Becker’s.
Which Hospitals Are Driving the Trend?
Not all hospitals are equally responsible. The study clearly shows that a small group is creating outsized impact. The top 10% of facilities drove most of the detected increases. This concentration matters because it points to specific adoption patterns rather than a system-wide problem.
AI Adoption Is Accelerating Rapidly
The broader context makes this trend urgent. According to a 2023 HFMA and AKASA survey, nearly half — 46% — of hospitals already use AI in billing, coding, and claims. Recent federal data shows the pace is quickening. By 2024, 7 in 10 U.S. hospitals used predictive AI. AI use for billing specifically jumped 25 percentage points year over year.
As AI shifts from experimental to essential, the stakes for accuracy are rising fast.
Industry Reactions and Pushback
Not everyone accepts the BCBS framing. Some hospital executives argue that coding has grown more complex for legitimate reasons.
The Hospital Industry’s Counterargument
Robert Boos, former chief revenue officer at Centra Health, previously argued that coding follows “strict federal and industry standards.” Health systems, he noted, invest heavily in compliance, training, and auditing.
The American Hospital Association (AHA) also offered context. The increasing complexity of inpatient admissions is partly because less-intensive care now moves to outpatient and office settings. As a result, what remains in inpatient settings is naturally higher acuity.
The Insurer Side Under Scrutiny Too
Notably, health insurers face their own AI controversies. UnitedHealth Group and Cigna are both facing lawsuits over their alleged use of algorithms to deny patients’ medical claims. This context is important. The billing dispute is not one-sided — it is increasingly an AI-versus-AI contest between hospitals and insurers.
What BCBSA Is Doing to Fix This
BCBSA is not simply raising alarms. The organization is actively working on solutions at both the national and local levels.
Key BCBSA Actions
- Establishing clear expectations for hospitals that use AI coding tools, so billing accurately reflects care delivered
- Harnessing data from over 60 million Americans to identify upcoding trends
- Collaborating with provider partners to better align payments with actual care
- Supporting BCBS companies with analysis and policy tools to counter aggressive coding
Dr. Razia Hashmi, BCBSA’s vice president of Clinical Affairs, summarized the concern. “Something is disconnected,” she said. “Among hospitals showing the fastest rise in diagnoses of post-partum anemia, the rise in patients coded with this condition wasn’t paired with the level of care we would have expected — and the patterns point to AI-enabled coding.”
Key Takeaways
- A new BCBSA/BHI study links hospital AI billing tools to over $2.3 billion in potential excess spending nationwide.
- The practice of upcoding — billing for conditions without matching treatment — is at the center of the findings.
- AI adoption in hospital billing is growing fast, with 7 in 10 hospitals using predictive AI by 2024.
- A small group of hospitals is responsible for the majority of coding intensity increases.
- BCBSA is working to increase transparency, establish oversight standards, and protect patients from inflated costs.
