Tucked into the latest budget reconciliation package for Congress, “One, Big, Beautiful Bill,” is a $25 million provision for the Department of Health and Human Services (HHS) to develop artificial intelligence (AI) tools aimed at reducing improper Medicare payments. According to the bill summary, the funding proposed in Sec. 112204 will support “tools for purposes of reducing and recouping improper payments under Medicare.”
Improper payments continue to be a persistent issue in both Medicare and Medicaid. According to the Centers for Medicare & Medicaid Services (CMS), Medicare’s traditional fee-for-service program recorded $31.7 billion in improper payments in fiscal year 2024, or 7.66% of total spending. Medicaid reported $31.1 billion in improper payments for the same period, representing 5.09% of the expenditure.
These improper payments include administrative errors and, in some cases, outright fraud. For example, Dr. Farid Fata, a Michigan oncologist, was convicted in 2014 and sentenced to federal prison for submitting $34 million in fraudulent Medicare and private insurance claims by administering unnecessary chemotherapy to patients without cancer.
States are also looking to AI for solutions to fraud. In January, Minnesota Gov. Tim Walz said the state would launch an anti-fraud initiative that includes AI tools for the state’s Medicaid billing. “As long as there have been programs aimed at helping people, there have been fraudulent actors looking to steal from those who need them most,” Walz said. “Our job is to stay one step ahead of them. We’re coupling new tools, like AI, with old-fashioned police work, to slam the door shut on theft.”
In a separate, early proposal from Congress, the bill Medicare Transaction Fraud Prevention Act (H.R. 7147) would establish a two-year pilot program to test predictive algorithms for identifying Medicare transactions that could be prone to improper payments (there was no budget attached to the pilot proposal).
Technology is advancing quickly. A 2024 study published in the Journal of Big Data tested AI tools on Medicare billing records and found that these tools improved the accuracy and clarity of fraud detection. Stella Batalama, dean of the Florida Atlantic University College of Engineering and Computer Science, which published the study, noted, “These methods, if properly applied to detect and stop Medicare insurance fraud, could substantially elevate the standard of health care service by reducing costs related to fraud.” The study used real Medicare data and showed that AI can help flag suspicious billing more efficiently than current approaches.
The move would advance HHS Secretary Robert F. Kennedy Jr.’s priority of using AI for agency efficiency. He has stated, “The AI revolution has arrived, and we are already using these new technologies to manage health care data more efficiently and securely.”
Evidence from recent studies and pilot programs suggests that AI has the potential to improve detection and prevention efforts, though the technology and its applications are still evolving. Continued investment in these tools may help agencies achieve more accurate oversight and reduce waste in public health care spending.