Blue Cross Blue Shield analysis ties hospital AI claims tools to $942M in extra spending

What the analysis found

Hospitals' use of artificial intelligence tools when submitting insurance claims added an estimated $942 million to healthcare spending over a two-year period, according to an analysis by the Blue Cross Blue Shield Association (BCBSA) reported by TechCrunch on September 26.

The insurer group's study identified a steep rise in patients documented with complex conditions. BCBSA argues there is a clear mismatch between that paperwork and reality, and found nothing to suggest the care those patients actually received changed accordingly.

Coding without matching care

At the center of the insurers' case is medical coding, the diagnostic shorthand providers use when billing for care. According to TechCrunch's account of the analysis, AI-assisted documentation is associated with more patients being coded as complex, while the treatment they received appears to have stayed the same.

That distinction matters because documentation of patient complexity feeds into payment. When coding shifts but care does not, added spending reflects paperwork rather than medicine — at least in the insurers' reading of the data.

AI on both sides of an old fight

The New York Times, cited by TechCrunch, treats the analysis as the latest sign that AI is pushing healthcare costs higher. Disputes between hospitals and insurers over treatment and payment are nothing new, the NYT noted, but the arrival of AI tools on both sides of those disputes appears to be making them worse.

That points to an escalating dynamic: providers adopt software to strengthen their claims, insurers respond with automation of their own, and each round raises the stakes for the next.

Voices from the industry

Two people quoted by TechCrunch frame the stakes differently. Dr. Shiv Rao, founder of AI healthcare startup Abridge, conceded that the technology could lead to a bleak future of automated systems contending with other automated systems over claims. He also argued the opposite is possible: that AI could reduce friction between hospitals and insurers and cut costs.

Luke Chalker, a senior vice president at BCBSA, rejected the idea that the situation is a two-sided fight. In his description, it is a one-sided rout, with insurers losing.

Why it matters

The BCBSA figure is one of the first attempts to quantify in dollars how AI is reshaping healthcare administration. It suggests AI in medical billing is not cost-neutral: software that improves documentation can also move large sums through the system without changing what care patients receive.

The analysis also comes from a trade group representing the paying side of the industry, so its framing deserves scrutiny. Hospitals are likely to argue that better coding simply captures the complexity of care they already deliver, while insurers read the same shift as inflated claims. How that argument resolves could shape regulation of AI-assisted billing, the pace of AI adoption in hospitals, and ultimately the premiums employers and patients pay.

There is a more optimistic scenario, and Rao named it: the same tools now raising costs could eventually standardize documentation and lower friction. Whether healthcare ends up with the bleak version or the efficient one will depend on how quickly providers, payers and regulators settle on rules for AI-generated claims.