BIS Examines Data, Tech, & Collaboration To Fight Money Laundering

New study from the BIS Innovation Hub finds that the current rules-based method of finding money laundering networks is less effective than using payment data, technologies that protect privacy, AI, and better cooperation to find money laundering networks.

The Financial Action Task Force says that almost all big plans to launder money are international and involve more than one type of business. Also, because their data and systems are broken up, it can be hard for financial institutions to find potentially suspicious networks and deals.

A Lexis Nexis study on the costs of financial crime compliance shows how much money financial institutions have to spend on AML. These prices went up by about $60 billion, to about $274 billion, between 2020 and 2022.

The Nordic Centre of the BIS Innovation Hub has been looking into new ways to solve the problem. Together with Lucinity, an Icelandic AI software-as-a-service business, they have been working on a proof of concept called Project Aurora.

The PoC used a large set of fake data that was made to look like real data on domestic and foreign payments. To protect private information, privacy-enhancing technologies were used, such as machine learning and other analytical tools, while the data remained encrypted.

Then, algorithms were taught on this set of fake data to find different patterns, called “typologies,” that are linked to money laundering across institutions and countries.

As part of the project, different ways of looking at the fake data were tested to represent different monitoring situations, such as siloed, national, and cross-border. Also, both at the national and cross-border levels, different ways to do collaborative research, such as centralized, decentralized, and hybrid models, were thought about.

BIS says that the results highlight the “effectiveness of employing advanced analytics and technologies that adopt a behavioural-based analysis approach, which focuses on understanding the relationships between different individuals and businesses and identifying anomalies from normal behaviour. The results demonstrated that such methods are more effective in detecting money laundering networks than are the current rules-based approach, which is limited by its siloed nature.”