Insurers: AI hospital claims add $942 million

Hospitals and insurers have long clashed over treatments and payments, but the New York Times pointed to the BCBSA analysis as the latest sign that AI on both sides is making things worse. The insurer group's data focuses on a documentation shift during claims submission: more patients recorded as having complex conditions, without what it calls evidence of a corresponding change in care delivered.

The disagreement over how to frame this is itself notable. Luke Chalker, a BCBSA senior vice president, rejected the language of a battle and described a completely one-sided blood bath with insurers losing. Dr. Shiv Rao of the AI startup Abridge offered the opposite possibility, warning of a horrible dystopic future where bots and agents fight each other, but also suggesting AI might reduce tensions and cut costs.

What happens next may hinge on whether payers can show that coding changes are untethered from actual treatment, or whether hospitals can demonstrate that AI is surfacing conditions that were previously missed. For insurers and the employers who fund coverage, the stakes are whether that $942 million figure reflects a one-time adjustment or an ongoing pattern, while for hospitals and clinicians the question is whether AI tools are treated as a billing problem or a care improvement.