Title: artificial intelligence: The New Secret Weapon in Uncle Sam’s Arsenal Against Fraud
Uncle Sam has recently introduced a new covert weapon to combat fraudsters attempting to pilfer funds from the American taxpayer: artificial intelligence (ai). Since late 2022, the Treasury Department has been utilizing advanced ai techniques to detect and prevent fraud. This strategy is a reflection of the private sector’s growing reliance on ai for rooting out suspicious transactions.
The Treasury Department’s use of ai-enhanced fraud detection has proven effective, recovering an impressive $375 million in fiscal 2023. This marks the first time the Treasury Department has publicly acknowledged employing ai for fraud detection purposes. With this new approach, the federal government can virtually halt check fraud in real-time by identifying abnormal transaction patterns. Additionally, the focus on ai has led to numerous active cases and arrests made by law enforcement.
It is important to clarify that the Treasury Department’s ai implementation does not involve generative ai, which generates content such as song lyrics or images from text prompts. Instead, their approach falls under machine learning and Big Data categories. The goal is to flag anomalies at such a rapid pace that banks are alerted before fraudulent checks are even cashed.
With the surge in fraud during the Covid-19 pandemic, the US government’s intervention is a welcome development. According to the US Government Accountability Office, as much as $135 billion in fraudulent Covid unemployment insurance claims were likely paid out by states. The Small Business Administration may have distributed over $200 billion in fraudulent Covid funds. Moreover, check fraud has increased by 385% since the pandemic. Financial institutions filed a staggering 680,000 suspicious activity reports (SARs) in 2022.
As one of the world’s biggest payers, Treasury’s mission to combat fraud is crucial, especially given its $6.9 trillion dispensation in 1.4 billion payments in 2021 alone, covering various categories from Social Security to tax refunds. Deputy Treasury Secretary Wally Adeyemo stated, “We are using the latest technological advances to enhance our fraud-detection process, and ai has allowed us to expedite the detection of fraud and recovery of tax dollars.”
The Internal Revenue Service (IRS) also started using ai in September 2021 to detect tax evasion, deploying “cutting-edge machine learning technology” on large and complex partnership tax returns.
Amiram Shachar, the co-founder and CEO of cloud Website security startup Upwind, advocates for the federal government’s use of ai to detect fraud. According to Shachar, “ai can help humans operate more efficiently, allowing them to see more in shorter timeframes.” In the public sector, where resource constraints limit human capacity, ai becomes an indispensable tool.
Financial institutions have been utilizing ai to verify transactions almost in real-time. Shachar shared that the technology can process large datasets with incredible speed, catching anomalies in milliseconds.
Mastercard’s recent announcement of a generative ai model to assist banks in identifying legitimate transactions marks an innovative leap in fraud detection. This ai technology is expected to analyze one trillion data points, potentially boosting fraud-detection rates by 20% or even up to 300%.
Despite the benefits of ai in combating fraud, concerns surrounding its potential risks to the financial system are growing. The Financial Stability Oversight Council, a team of leading regulators led by the Treasury secretary, designated ai as an “emerging vulnerability” late last year. Experts are particularly concerned about how ai could supercharge financial fraud. A recent case of a finance worker being tricked into transferring $25 million to fraudsters using deepfake technology underscores these concerns.
