TLDR:
- Treasury Dept used AI to recover $1 billion in check fraud in FY 2024
- Total fraud prevention/recovery reached $4 billion, up 6x from previous year
- Machine learning AI analyzes data to detect fraud patterns quickly
- AI helps protect $7 trillion in annual Treasury payments
- Human oversight remains key in final fraud determinations
The U.S. Department of Treasury has made strides in combating financial crime through the use of artificial intelligence. In fiscal year 2024, the department recovered $1 billion in check fraud alone, nearly tripling the amount from the previous year.
This success is part of a larger effort that prevented or recovered more than $4 billion in fraud overall, marking a six-fold increase from fiscal year 2023.
The Treasury’s fraud detection initiative, which began in late 2022, utilizes machine learning AI to analyze vast amounts of data and identify suspicious patterns. This technology allows officials to sift through information much faster than human analysts, detecting anomalies and potential fraud in milliseconds.
Renata Miskell, a senior Treasury official, described the impact of AI on their operations as “transformative.” The department is responsible for distributing approximately 1.4 billion payments annually, totaling nearly $7 trillion, to around 100 million recipients. This massive volume of transactions makes the Treasury a prime target for fraudsters, necessitating advanced detection methods.
The AI systems employed by the Treasury focus on machine learning rather than generative AI. These algorithms excel at analyzing data streams and making rapid decisions based on learned patterns.
While the technology flags suspicious transactions, Miskell emphasized that human oversight remains crucial, with federal agencies making the final determination on fraud cases.
The Treasury’s adoption of AI for fraud detection follows similar practices already in use by many banks and credit card companies. The department is exploring further enhancements to its fraud-detection tools, including testing new data sources and collaborating with state agencies to combat unemployment insurance fraud.
The use of AI in financial crime prevention has become increasingly important as online payment fraud continues to grow. Juniper Research estimates that such fraud could exceed $362 billion by 2028. However, the technology itself presents new challenges, as demonstrated by recent incidents involving deepfake videos used in elaborate scams.
Treasury Secretary Janet Yellen has acknowledged both the potential and risks associated with AI in finance.
In June, she warned bankers about the “significant risks” posed by AI, and regulators have classified it as an “emerging vulnerability” to the financial system.