AI Helped the Feds Catch $1 Billion of Fraud in One Year. And It’s Just Getting Started
The US federal government’s strategic adoption of artificial intelligence (AI) to combat financial crime is yielding remarkable results. In fiscal year 2024, machine learning-powered AI enabled the US Treasury Department to identify and recover $1 billion in check fraud, a significant increase from the prior year’s recovery totals.
“It’s really been transformative,” said Renata Miskell, a senior Treasury official. “Leveraging data has upped our game in fraud detection and prevention.”
In total, AI-assisted efforts helped the Treasury recover and prevent more than $4 billion in fraud in fiscal 2024—a sixfold increase from the year before. This milestone underscores AI’s potential to protect taxpayer funds and improve the efficiency of fraud detection processes.
AI’s Role in Combating Fraud
The Treasury began implementing AI for financial crime detection in late 2022, adopting techniques long used by banks and credit card companies. Unlike generative AI tools such as ChatGPT, these efforts rely on machine learning, a branch of AI adept at analyzing massive datasets and identifying patterns.
Machine learning models can analyze vast amounts of data and detect subtle anomalies in a fraction of the time it would take humans. Once trained, these sophisticated models flag suspicious transactions in milliseconds, enabling rapid intervention.
“Fraudsters are really good at hiding. They’re trying to secretly game the system,” Miskell explained. “AI and leveraging data helps us find those hidden patterns and anomalies and work to prevent them.”
This capability is vital for the Treasury, which processes approximately 1.4 billion payments annually, valued at nearly $7 trillion. These payments encompass Social Security benefits, Medicaid payments, tax refunds, and more, making the department a lucrative target for fraudsters.
Addressing New Challenges
While AI strengthens fraud prevention efforts, it also introduces new vulnerabilities. Fraudsters increasingly leverage AI to create sophisticated scams, including deepfake videos and synthetic identities. For example, a finance worker in Hong Kong was tricked into transferring $25 million to fraudsters using a deepfake video earlier this year.
Recognizing these risks, Treasury Secretary Janet Yellen has classified AI as an “emerging vulnerability” in the financial system. However, Miskell emphasized that human oversight remains integral to AI-driven fraud detection. “A human is always in the loop,” she noted, ensuring that final determinations are made by federal agencies.
Expanding AI’s Reach
The Treasury’s use of AI is still evolving. Officials are exploring advanced fraud-detection methods similar to those employed by leading financial institutions. This includes testing new data sources and collaborating with state agencies to combat unemployment insurance fraud.
The stakes are high: online payment fraud is projected to exceed $362 billion by 2028, according to Juniper Research. By enhancing its AI capabilities, the Treasury aims to stay ahead of increasingly sophisticated fraud tactics.
Looking Forward
The success of AI in combating fraud demonstrates its transformative potential in securing taxpayer money. As the Treasury continues to refine its tools and collaborate with other agencies, AI’s role in financial crime prevention will likely expand, creating a more resilient system against emerging threats.
With fraud detection powered by cutting-edge AI, the federal government is not only safeguarding public funds but also setting a new standard for leveraging technology to combat crime on a national scale.