AI-Driven Coding Inflates Medical Bills by Nearly $1 Billion

The Blue Cross Blue Shield Association (BCBSA) found that the proportion of inpatient cases classified as medically complex climbed from 37% in early 2023 to 40% by the end of 2025. This shift resulted in $942 million in additional costs over two years, with $653 million linked directly to secondary diagnoses that moved claims into higher payment brackets. Luke Chalker, BCBSA’s senior vice president of product and data science, noted that the data shows no corresponding change in actual patient care.

This trend highlights a growing friction between hospitals aiming for maximum revenue and insurers attempting to curb payouts, with AI acting as the primary lever. While hospitals utilize AI to scan records and notes for billable conditions, the lack of increased treatment—such as blood transfusions for patients newly diagnosed with anemia—suggests that the technology is optimized for profitability rather than clinical necessity. Previous BCBSA findings in maternity care mirror this pattern, where diagnosis rates for posthemorrhagic anemia spiked while medical interventions remained stagnant.

Progressive lawmakers are using these figures to argue that current market incentives are fundamentally flawed. Senators like Bernie Sanders and Representative Pramila Jayapal contend that a for-profit insurance model cannot be corrected through minor adjustments. They advocate for a universal public program, citing research that suggests a Medicare for All system could significantly reduce national healthcare spending while improving health outcomes.