A new Blue Cross Blue Shield Association analysis suggests artificial intelligence billing software may be pushing up hospital costs even when patients are not actually sicker and their treatment is no more intensive.
The association said the tools increased healthcare spending for Blue Cross and Blue Shield companies by an estimated $942 million from 2023 to 2025, according to Quartz.
Here's what to know
A claim can be moved into a higher-reimbursement category when software identifies secondary diagnoses, meaning conditions beyond the main reason for a patient's visit. The analysis said more than 60% of hospital systems now use AI tools to scan lab results and electronic health records for those additional diagnoses, even when the care provided does not change.
Most of the increase the association tracked came from secondary diagnoses, which researchers said produced about 70% of the added spending — more than $650 million over the two-year period. That estimate came from de-identified claims data from member companies that cover one in three Americans, with the analysis centered on inpatient settings.
Luke Chalker, the association's senior vice president of product and data science, said the core issue is the disconnect between coding and care.
"If patients are truly sicker, we'd expect to see more treatment," Chalker said. "The disconnect between diagnoses and treatment suggests that AI is identifying more billable conditions, not sicker patients."
Looking at major bowel procedures, researchers found diagnoses becoming more common without a matching rise in treatment. From the first quarter of 2023 to the fourth quarter of 2025, diagnoses such as partial intestinal blockages rose 55% and acid overload rose 33%, while treatment rates for those conditions stayed flat.
More background
Hospital billing is not limited to administrative paperwork. When insurers pay more, those costs can show up in premiums, employer healthcare spending, and out-of-pocket bills for families already struggling with affordability.
One cost-raising pathway described in the report starts with a single lab result. Software scanning a patient's record can flag a secondary diagnosis from that value and move the claim to a more severe reimbursement level even though the patient's care stays the same.
What can be done?
By publishing its analysis, the Blue Cross Blue Shield Association is calling for closer scrutiny of how hospitals use coding software and whether reimbursement rules need stronger guardrails when AI enters the billing process.
Patients can review explanations of benefits, request itemized hospital bills, and ask for clarification when a diagnosis appears to have changed the price of a stay.
Employers and insurers, meanwhile, may increasingly turn to audits that compare diagnoses with actual treatment patterns. That kind of review could help distinguish between more accurate documentation and coding changes that simply generate larger bills.
"As families face higher healthcare costs, this research underscores the urgent need to better understand these AI tools and the role they may play in exacerbating the affordability crisis," said David Merritt, the association's senior vice president of external affairs.
Where can I learn more?
Questions about AI in hospital billing are part of a wider shift across healthcare, from insurance decisions and diagnostic tools to the infrastructure needed to run these systems.
• In California, lawmakers pushed to curb insurance companies from denying claims with AI alone.
• Researchers found very small words make AI health chatbots change medical advice unpredictably.
• Clinicians deployed a groundbreaking AI test to detect tuberculosis much faster at a health center.
• Scientists warned that AI-related data center pollution is emerging as a public health threat.
• Utilities are signing unprecedented deals with this industry as AI data centers strain power grids.
Get TCD's free newsletters for easy tips, smart advice, and a chance to earn $5,000 toward home upgrades. To see more stories like this one, change your Google preferences here.







