Insurers blame hospitals' AI coding for $942 million in extra claims
The Blue Cross Blue Shield Association says AI billing tools are inflating hospital bills. The hospital lobby says the analysis is bunk. Also today: Microsoft quietly retires the Copilot+ PC brand.
The Blue Cross Blue Shield Association says hospitals' AI coding tools added $942 million to its members' costs over two years — and the hospital lobby calls the finding unsubstantiated.
The association's September 24 white paper, built on an analysis of major bowel procedures, found hospitals classified 55,158 additional cases as complex between early 2023 and late 2025, generating $653 million in incremental reimbursements at an average of $11,800 per excess case. The share of cases billed as complex rose from roughly 37% in early 2023 to about 40% by late 2025, and secondary diagnoses accounted for about 70% of the added inpatient coding cost. Claims coded as the most complex rose from 20.2% to 22.7% for that procedure family alone, worth $60.8 million.
More than 60% of hospital systems have started using AI coding tools, BCBSA says, drawing on claims data from the 31 independent Blue plans that cover more than 100 million people. The association's Luke Chalker argues the new diagnoses aren't matched by treatment: ICU use, transfusions and median length of stay were flat or lower at the hospitals whose complex-case rates grew fastest. "The disconnect between diagnoses and treatment suggests that AI is identifying more billable conditions, not sicker patients," he said.
Hospitals have a blunt answer. McLaren Health Care's chief financial officer told the New York Times his system buys AI chart-review software from SmarterDx and has added about $1 million a month in revenue — while insurers "are using A.I. to scan our charts to look for claims to deny." The American Hospital Association calls the upcoding claim "unsubstantiated," pointing to a roughly 5% rise in its case-mix index from sicker patients. BCBSA's own paper concedes the limits: it uses claims rather than clinical records, so it cannot verify whether any individual patient was sicker, and it names no hospital and no vendor. We've covered AI pressure on public claims before — AI agents flood public services — and the claims are mostly real — but this is the first time the money is quantified in the billions.
Microsoft has quietly retired the "Copilot+ PC" label, two and a half years after launching it as the defining category of AI PCs.
Surface corporate vice president Brett Ostrum told Windows Central that the new 12-inch Surface Pro and 13-inch Surface Laptop, which still meet the old spec bar, are "not called Copilot+ PCs." Qualcomm's Kedar Kondap said devices delivering the same experiences will "probably" drop the terminology too. The bar was always going to be outgrown — a 45 TOPS NPU requirement that the wider PC market has now reached — and the brand's flagship feature, Recall, shipped late after security researchers tore it apart. Nvidia never used the label for its RTX Spark machines, which meet the spec. No replacement name has been announced; Microsoft holds a Windows and Surface event on October 7 with Satya Nadella, Pavan Davuluri and Jensen Huang, billed as a conversation on how local AI shapes the next chapter of the PC.
What to watch: whether October 7 names a successor brand — and whether the coding fight between hospitals and insurers ends up in front of a regulator instead of a claims auditor.
Should insurers be allowed to audit the AI tools hospitals bill with, or does that just start a bot war over every chart? Tell us in the comments.
