AI Helps U.S. Treasury Recover $4 Billion in Fraud Prevention Efforts

The U.S. Treasury Department’s use of artificial intelligence to combat financial crime is showing significant results. According to new estimates shared with CNN, the department leveraged machine learning AI to recover $1 billion in check fraud in fiscal year 2024 alone, nearly tripling the amount recovered in the previous year.

AI’s Role in Detecting Fraud

AI has played a key role in helping the Treasury detect and prevent financial fraud, with overall recovery reaching $4 billion in 2024, a six-fold increase from the prior year. According to Treasury official Renata Miskell, AI has "transformed" the department’s ability to sift through vast amounts of data to uncover fraud, a task that would have been much slower and less accurate with human-only methods.

“Leveraging data has upped our game in fraud detection and prevention,” Miskell said in an interview with CNN.

How AI Works in Fraud Detection

Unlike generative AI tools like OpenAI’s ChatGPT or Google’s Gemini, which create content, the Treasury’s fraud detection relies on machine learning, a subset of AI that excels at data analysis. Machine learning models can comb through enormous datasets, detect subtle patterns, and identify suspicious transactions in milliseconds. This capability is crucial in the Treasury’s efforts to protect taxpayer money, especially after fraud spiked during the COVID-19 pandemic.

“Fraudsters are really good at hiding,” Miskell noted. “AI and leveraging data help us find those hidden patterns and anomalies and work to prevent them.”

Treasury’s Expanding Role in Fraud Prevention

The U.S. Treasury is responsible for handling approximately 1.4 billion payments annually, valued at nearly $7 trillion, which makes it a prime target for fraudsters. From Social Security and Medicaid payments to federal paychecks, tax refunds, and stimulus checks, the Treasury delivers payments to over 100 million people each year. The sheer volume of transactions makes AI a valuable tool in the fight against fraud.

The Treasury is now expanding its AI efforts, including testing new data sources and collaborating with state agencies to tackle unemployment insurance fraud. Miskell hinted at future plans to incorporate fraud detection methods used by banks and credit card companies but declined to share details to avoid tipping off potential criminals.

Challenges and Concerns Over AI in Finance

While AI is proving effective in preventing fraud, it also poses risks to the financial system. Treasury Secretary Janet Yellen warned in June that AI could introduce new vulnerabilities in finance, and top regulators classified AI as an “emerging vulnerability” to the financial system. Some of these concerns stem from AI’s potential to be exploited by criminals. For instance, deepfake technology has already been used in fraud schemes, such as the infamous case where a finance worker in Hong Kong was tricked into transferring $25 million to fraudsters.

Despite these risks, Treasury officials remain confident in AI’s ability to detect fraud, emphasizing that human oversight is always involved in determining whether a flagged transaction is fraudulent.

The Future of AI in Fraud Detection

The Treasury’s use of AI to fight financial crime is still in its early stages, but the results so far indicate a promising future. With plans to speed up the development of fraud-detection tools and expand collaboration with state agencies, AI will likely continue to play an increasingly critical role in protecting taxpayer dollars from fraud.