On October 18th, the US Department of Treasury announced that their latest effort in preventing fraud and improper payments has recovered over $4 B which was lost in the year.
The department shared that they took a technology and data-driven approach. The recovery of funds lost in fraudulent payments has increased by $652 million as compared to last year.
In a report, the Treasury Department’s Office of Payment Integrity, which is within the Bureau of the Fiscal Service, stated that it had started to use “machine learning AI” to detect high-risk transactions. These tools were instrumental in averting $2.5 billion of fraudulent transactions. Also, the Treasury’s check fraud detection systems helped to prevent $1 billion of fraud.
The use of AI has been instrumental in the fight against fraud. Wally Adeyemo, Deputy Secretary of the Treasury, highlighted the role of these efforts in a press release, stating, “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.”
AI used across Federal agencies to tackle fraud
In addition to fraud prevention, the Treasury also made changes to the payment processing schedules, which saved $180 million in improper payments. These measures also enhanced risk-based screening mechanisms and prevented $500 million of fraud. As a result, the department has improved its capacity to distinguish between fraud and genuine claims, particularly on payments, thus saving taxpayers money.
Furthermore, to promote the access and usage of the department’s payment integrity solutions, the Treasury will focus on establishing partnerships with new and high-risk programs. In May, the Treasury announced it’s partnership with the Department of Labor.
The application of AI is not limited to the Treasury Department only. Other government departments, like the Internal Revenue Service (IRS), also use AI to fight financial crimes. The IRS started using AI in September 2023 to catch tax evaders by scrutinizing tax returns filed by big taxpayers and other legal entities, including hedge funds and law firms.
Moreover, the National Oceanic and Atmospheric Administration (NOAA) started using neural networks in 2024 to improve hurricane forecasting and to provide warnings. These initiatives’ success shows how AI is increasingly being used in different sectors of the federal government.
Currently, the Treasury handles 1.4 billion transactions per year, totalling $6.9 trillion. In response to the increasing cases of fraud, the agency unveiled plans in May 2024 to boost its AI technology in order to strengthen the surveillance and enforcement of financial rules against fraud. Notably, Juniper Research recently reported that online payment fraud is likely to grow past $362 billion by 2028.