Artificial intelligence tools used by hospitals for documentation and coding of services are estimated to have added US$ 942 million to Blue Cross Blue Shield plans' spending between 2023 and 2025, according to an analysis by the Blue Cross Blue Shield Association. The association identified an increase in the classification of patients as clinically complex cases without a proportional increase in treatment intensity.
The estimate was calculated from de-identified claims data from the companies that make up the association. The share of hospitalizations classified as complex rose from about 37% at the beginning of 2023 to 40% at the end of 2025.
The increase matters because complexity classification influences the amount paid to hospitals. AI tools can analyze medical records and other clinical data to identify additional diagnoses capable of placing a hospitalization in higher-reimbursement categories.
Secondary diagnoses accounted for US$ 653 million
According to BCBSA, approximately US$ 653 million, or 70% of the estimated additional cost, came from more than 55,000 cases in which secondary diagnoses elevated hospitalizations into categories of higher severity and payment.
The association says it did not find an equivalent change in care in those cases. At the same time, it emphasizes that the US$ 942 million is an estimate based on the patterns observed in the claims, and not a demonstration that every adoption of AI increases medical expenses.
Hospitals dispute the interpretation that greater coding intensity necessarily represents excessive billing and argue that the technology can also improve documentation accuracy.
The dispute comes as automation advances on both sides: hospitals use AI to identify codes and support reimbursement requests, while insurers apply similar systems to review claims and challenge charges.
