Hospital AI Use Tied to Nearly $1 Billion in Extra Insurance Payouts, Blue Cross Group Says

BCBSA found that using AI to scan medical records and transcribe patient meetings may be contributing to rising costs. The increase is not mainly due to additional computing power, the report said, but to AI’s ability to scan large volumes of current and old data and uncover secondary ailments and complications. That can lead to more diagnoses being recorded and, in the US reimbursement system, more payouts.

A patient may enter the hospital for one primary problem, but may also have several secondary diagnoses or complications. Because AI is effective at scanning data, records and conversations, it is more likely to identify those conditions than a human reviewer. US hospital reimbursement is influenced by diagnoses and complexity, so additional diagnoses can mean additional payments.

The association warned that if patients were really becoming sicker, treatment increases should be visible. They are not, according to the report. “The disconnect between diagnoses and treatment suggests that AI is identifying more billable conditions, not sicker patients,” said Luke Chalker, SVP of Product and Data Science at BCBSA.

Insurers are also increasingly using AI to scrutinize claims and decide whether treatment was medically necessary. TechRadar described this as an AI-vs-AI battle, with both providers and insurers using AI to analyze records, claims and clinical information. That can push up compute consumption and costs across the board, the report said, potentially without a corresponding health benefit.

The BCBSA represents 31 independent Blue Cross Blue Shield insurers and more than 100 million citizens. Its estimate covers 2024-2025 and compares payouts with the prior year. The report said the cost pressure comes from AI-assisted hospital documentation and other AI work, and that insurers are using AI to question whether diagnoses are relevant.

Editor's Summary

The Blue Cross Blue Shield Association says AI-assisted hospital documentation and other AI work led insurers to pay nearly $1 billion more than usual in 2024-2025. BCBSA links the increase to AI finding more billable secondary conditions without evidence that patients are sicker. Insurers are also using AI to review claims, creating an AI-vs-AI dynamic that may raise system costs and premiums.