- Insurers paid out $1bn extra because of AI assistance in hospitals, report claims
- AI is scanning through swathes of current and old data to find extra ailments
- Insurers are using AI to question whether those diagnoses are relevant
The Blue Cross Blue Shield Association (BCBSA) has revealed that insurers paid out nearly $1 billion more than usual from AI-assisted hospital documentation and other AI work in 2024-2025, compared with the year before.
The organization, which represents 31 independent Blue Cross Blue Shield insurers (and over 100 million citizens), found that using AI to scan medical records and transcribe patient meetings could actually be causing costs to rise, and even be passed on to patients via higher premiums.
But the cost isn't coming from the extra compute – that's a whole other issue. Instead, it's AI's ability to scan through huge amounts of data that could be revealing secondary ailments and pushing insurance payouts up.
AI is uncovering secondary ailments and pushing up insurance payouts
While a patient may enter the hospital for one primary problem, they may also have several secondary diagnoses or complications. Because AI is so effective at scanning through swathes of data, records and conversations, it's more likely to pick these up than a human reviewer. But while this is ultimately good news for the patient's health, US hospital reimbursement is influenced by diagnoses and complexity, so more diagnoses means more payouts.
However, the BCBSA warns that if patients really were becoming sicker, we should be seeing increases in treatment – but we're not. "The disconnect between diagnoses and treatment suggests that AI is identifying more billable conditions, not sicker patients," SVP of Product and Data Science Luke Chalker said.
But on the other side of the coin, insurers are also increasingly employing AI to scrutinize claims and decide whether treatment was medically necessary, making this somewhat of an AI-vs-AI battle, pushing up compute consumption and costs across the board and potentially for no beneficial reason.
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds.
