The Blue Cross Blue Shield Association says AI coding tools have added $942 million in hospital costs over two years, with no evidence patients got any sicker. The finding comes from a BCBSA claims analysis reported by TechCrunch and Fierce Healthcare.
The share of inpatient stays billed as medically complex rose from 37% in early 2023 to 40% by the end of 2025, BCBSA found. About $653 million of the added cost traces to secondary diagnoses that pushed claims into higher-paying diagnosis-related groups, adding roughly $11,000 per case, and more than 55,000 additional complex cases were billed against the 2023 baseline, Fierce Healthcare reported.
BCBSA compared coding complexity against actual treatment intensity, including ICU use, transfusions, reoperations and length of stay, and found hospitals with the sharpest jumps in complexity billing showed similar or lower treatment intensity than their peers. Coding "has materially changed" without a corresponding shift in care delivered, said Luke Chalker, BCBSA's senior vice president, and Razia Hashmi, the association's vice president of clinical affairs, said "the likelihood that this is technology-enabled upcoding is higher," according to Fierce Healthcare.
Chalker described the dispute between insurers and hospitals over AI-driven coding as "not a war," calling it "a completely one-sided blood bath" with insurers on the losing end, TechCrunch reported. Dr. Shiv Rao, founder of AI documentation company Abridge, acknowledged the risk but said AI could also ease tensions and cut costs over time, according to TechCrunch.
For anyone selling AI into health care, or building on hospital data, this is a preview of the backlash: a payer with its own claims data can now point to a specific dollar figure and pin it on a specific category of tool, not a vague unease about AI.