The US Department of the Treasury employed artificial intelligence to analyze data and recover over $4 billion obtained through fraud and misconduct in the 2024 fiscal year.
Through the use of AI, the department managed to reclaim six times more funds compared to the previous reporting period, when the amount was $652.7 million, the Treasury noted.
The ministry provided a detailed account of the results achieved through the application of neural networks:
- expansion of risk-based screening prevented criminal transactions worth $500 million;
- identification and prioritization of high-risk operations accounted for $2.5 billion;
- accelerated detection of fraud involving treasury checks led to the recovery of $1 billion;
- enhanced efficiency in payment processing schedules halted fraud amounting to $180 million.
At the end of 2022, the Treasury began utilizing machine learning based on artificial intelligence—a technology for analyzing vast amounts of data, forecasting, and decision-making based on the information obtained.
“Fraudsters are indeed adept at hiding. They try to deceive the system. AI and data usage help uncover hidden patterns and anomalies and work on preventing them,” remarked Treasury representative Renata Miskell.
In August, Binance prevented the theft of over $2.4 billion in user funds using AI.