Photo credit: Sven Hansche / Shutterstock.com
The Bank of England (BoE) has tested the use of artificial intelligence in both real-time investment fraud detection and inflation forecasting.
In partnership with the BoE, the Bank for International Settlements (BIS) Innovation Hub London Centre has published the results of Project Hertha, which tests the “application of modern AI techniques to help spot complex and coordinated criminal activity in payment system data”.
The trial used AI-enabled models to identify complex patterns as well as a synthetic dataset comprising 1.8 million accounts and over 308 million transactions over the course of a year. The results indicate that findings from payment system analytics helped banks and payment service providers detect 12% more illicit accounts, alongside a 26% improvement in spotting new criminal patterns.
The project highlights the potential of AI to enhance fraud detection by detecting complex networks and behaviors. Although the study acknowledges that the results are promising, it argues “there are limits to the application and effectiveness of system analytics”.
In addition to fraud detection, the BoE is exploring how AI could help in forecasting inflation. BoE Monetary Policy Committee member Megan Greene said AI could help produce more accurate projections and enhance policy communication. AI could also provide early warning signs of a financial crisis and help analyze labor markets.
Both of these initiatives reflect the increasing use of AI in an effort to enhance financial stability and policy-making efficacy.