AI Could Uncover Hidden Medical Bill Discounts for Patients but Experts Warn It May Drive up Healthcare Costs

Patients in the US are increasingly turning to artificial intelligence to challenge hospital bills, identify potential errors and uncover discounts. But experts warn that the technology could also drive up healthcare spending and expose consumers to financial and privacy risks.

Tools such as ChatGPT and Claude can review medical statements, explain billing codes and compare invoices with Explanation of Benefits documents.

They can flag possible duplicate charges, upcoding and unbundled services, helping patients question bills without specialist knowledge of insurance rules. However, these tools are not a substitute for professional legal or medical advice.

Some Patients Report Significant Reductions in Medical Bills

In a case shared publicly on Threads by a relative, a family said it used AI to analyse a $195,628 hospital bill and prepare a dispute. The family reported that the balance fell by about $163,000, leaving roughly $33,000 to pay.

Healthcare platforms such as Cedar say they use hospital and physician data to predict patients' willingness to pay and select discounts. They can then offer discounts or interest-free payment plans to people facing substantial medical debt.

A broader concern is that similar predictive models, if used by debt collectors or debt buyers, could help determine how aggressively to pursue unpaid balances. Critics say that, without transparent oversight, such technology could prioritise revenue recovery over meaningful relief for patients.

AI-Assisted Coding Linked to Higher Healthcare Spending

A Blue Cross Blue Shield Association analysis, carried out by its licensee Blue Health Intelligence, estimated that hospitals' use of AI-assisted coding may have contributed to $942 million in additional costs across 2024 and 2025, compared with a 2023 baseline.

The analysis said AI scribes generate clinical documentation, after which coding software can recommend more intensive service codes. This can classify patients as sicker without a corresponding increase in treatment.

The analysis pointed to rising diagnosis codes for acute posthaemorrhagic anaemia, a condition associated with severe blood loss, without a matching increase in blood transfusions.

The American Hospital Association disputes allegations of inappropriate spending. It argues that rising coding reflects more precise documentation and a sicker patient mix, rather than improper charges.

Analysts have described a billing 'arms race', with health systems using AI to identify billable conditions while insurers deploy their own systems to challenge or deny claims. Mehmet Oz, who leads the Centers for Medicare and Medicaid Services, and industry analysts have warned that this competition could increase costs before any long-term savings emerge.

An AI Review Can Also Increase a Hospital Bill

In a case described in a TikTok video by a social worker, a patient challenged an $85,000 hospital bill. The social worker said the hospital subsequently used an 'AI sweep' to identify nearly $20,000 in additional charges, raising the total to $102,000.

According to the social worker, the hospital said the charges had been omitted from the original bill. Neither the hospital's explanation nor the reported use of AI has been independently confirmed. The social worker argued that patients should be able to question bills without fearing a higher charge as a result.

Consumer AI tools generally lack access to patients' full medical records, physicians' notes and insurance contracts. They cannot reliably determine whether a charge is justified or legally incorrect, and may flag legitimate charges as errors.