Treasury Department uses AI to recover $1 billion in fraud

The Treasury Department is taking advantage of artificial intelligence (AI) to prevent or recover check fraud and stopped more than $1 billion in fraudulent payments over the past year, the agency announced Thursday.

Machine learning AI was used by the Treasury Department to recover about $1 billion in check fraud in fiscal 2024, the agency said in a press release. The increased use of AI was part of various enhanced fraud detection processes the department rolled out over the fiscal year. 

The enhanced measures helped prevent or recover more than $4 billion total in fraud, according to the agency.

The machine-learning algorithms processed and learned with mass amounts of data to help the Treasury Department more easily flag potential fraud schemes. The department declined to share specific examples, citing the nature of the operations.

The total amount of fraud recovered or prevented also included $2.5 billion by identifying and preventing high-risk transactions, $500 million through expanded “risk-based” screenings and $180 million through increasing the efficiency of payment processing schedules. 

“Treasury takes seriously our responsibility to serve as effective stewards of taxpayer money. Helping ensure that agencies pay the right person, in the right amount, at the right time is central to our efforts,” Deputy Treasury Secretary Wally Adeyemo said in a statement.  

Online payment fraud is expected to surpass $362 billion by 2028, the Treasury Department stated, citing findings from Juniper Research.  

The Treasury Department’s embrace of AI follows a similar move by the IRS, which launched an initiative last year using the new technology to crack down on back taxes. In this case, AI technology can discover patterns of tax avoidance more easily than before.