Banks Use AI to Block Trillions in Terrorism and Trafficking Funds

AI is increasingly being used to analyze financial data for signs of criminal activity, helping to curb the flow of illicit funds worldwide.

Financial institutions and regulatory technology (RegTech) firms are harnessing artificial intelligence to intercept the trillions of dollars channeled into illegal activities. According to Nasdaq’s latest Global Financial Crime Report, approximately $3.1 trillion in illicit funds moved through the global financial system last year. This included substantial amounts for money laundering, with $346.7 billion linked to human trafficking, $782.9 billion to drug trafficking, and $11.5 billion to terrorist financing.

As financial criminals become more sophisticated, leveraging advanced technologies that are increasingly accessible and affordable, financial institutions and RegTech companies are employing similar technologies—such as AI and generative AI—to combat these threats. 

Nikhil Aggarwal, managing director at Deloitte Transactions and Business Analytics, explains, “AI enables us to see how bad actors interact with each other. By visualizing broader networks, we can conduct deeper investigations into criminal rings and uncover interconnected patterns among these threat actors.”

The U.S. Bank Secrecy Act, enacted in 1970, aimed to help financial institutions detect and prevent money laundering, also known as Anti-Money Laundering (AML) laws. This includes the KYC (Know Your Customer) process, which financial institutions use to verify the identity and assess the risks of potential clients. 

However, with the rise of digital transactions—such as online payments and deposits—financial crime has surged. More than half of Americans now use digital wallets more frequently than cards or cash, according to a Forbes Advisor poll from last year. This has generated vast amounts of transaction and customer behavior data, which AI technology can now analyze more effectively.

Dagan Osovlansky, chief product officer at Israeli software company ThetaRay, notes, “AI excels at analyzing large datasets and identifying patterns on a massive scale. Many bankers have extensive data but often lack the means to use it effectively.”

ThetaRay, utilizing its proprietary machine learning algorithms, applies a risk-based approach to financial crime detection. Its AI learns normal banking behaviors through “unsupervised learning”—a method that identifies anomalies without human oversight. The company’s platform monitors over $15 trillion in transactions and recently acquired Screena, an AI-powered screening firm, to enhance its capabilities.

Despite these advancements, Osovlansky suggests that criminals may still be outpacing financial institutions in adopting new technologies, stating, “The real question is, who’s winning? I’m not sure the answer is favorable for us. I believe we are playing catch-up.”

Early implementations of AI in financial crime detection have already reduced false positives—instances where normal activity is mistakenly flagged as suspicious—at several partner banks, including Santander, which has used ThetaRay’s solution since late 2019.

Other RegTech players, such as Iceland-based Lucinity, also leverage AI to enhance financial crime compliance. Meanwhile, some institutions, like HSBC, have developed their own AI systems in collaboration with Google to monitor 1.2 billion transactions across 40 million accounts each month, reportedly spotting two to four times more financial crime with 60% fewer false positives.

JPMorgan Chase, another leader in AI adoption, plans to make its KYC processes, including customer onboarding and monitoring, up to 90% faster by the end of next year. This efficiency is expected to streamline the processing of 230,000 files with 20% fewer staff.

The main challenge for financial institutions and their partners remains data quality and availability. According to Deloitte’s Aggarwal, improving data cleanliness and integration is crucial for unlocking more powerful insights from AI. “Addressing data fundamentals is where the real opportunity lies,” he concludes.