An insurance industry analysis has said the spread of hospitals' AI medical coding tools could lead to higher U.S. healthcare costs. It said more cases are being classified as more complex conditions, resulting in about $942 million in additional spending over two years.

On Sept. 26 (local time), IT outlet TechCrunch reported that the Blue Cross Blue Shield Association (BCBSA) said its analysis of claims data from 2023 to 2025 showed a sharp rise in cases in which hospitals classify patients as medically more complex. It added that no corresponding change was confirmed in the treatment actually provided.

More than $650 million, about 70 percent of the additional costs, was linked to secondary diagnoses added beyond the main reason for admission. AI can identify such conditions by analysing test results and medical records, and additional diagnoses can move patients into billing groups with higher severity levels and reimbursement rates. BCBSA said more than 60 percent of U.S. hospital systems are using AI-based coding technology.

The analysis alone makes it difficult to conclude that AI is a direct cause of higher medical costs. Citing a link between the spread of AI coding tools and rising billing complexity, BCBSA raised the possibility that AI may be contributing to higher costs.

Concerns have also emerged that a new automation race could unfold as both hospitals and insurers use AI. Shiv Rao (시브 라오), founder of Abridge (에이브리지), warned, "In the future, bots could fight bots, and agents could fight agents," but said, "If AI is used properly, it could instead lower administrative friction and costs."

As AI draws attention as a technology to improve the efficiency of medical administration, the impact of diagnosis and claims automation on actual medical costs is emerging as a new issue.