The feds accelerate fraud detection with machine learning, catching up to similar techniques already in use by leading banks and credit cards.
The US Treasury Department says AI tools prevented criminals from stealing $4 billion in taxpayer dollars this year, a dramatic increase from $653 million last year.
The tools have been “transformative”, Renata Miskell, a top Treasury official, tells CNN. They don’t use generative AI, like ChatGPT, but rather are machine learning-based. They’re helping identify high-risk transactions, expedite check fraud, implement efficiencies, and expand risk-based screenings.
The Treasury is a prime target for financial fraud because it doles out huge sums of cash every year for Social Security, Medicaid, tax refunds, federal paychecks, and stimulus checks.
