It's a lot easier for banks to crack down on money launderers when they can share data among themselves. Because of privacy rules, however, this is only possible to a limited extent. Together with TNO, Rabobank and ABN AMRO have now developed what is known as multi-party computation. This allows them to jointly analyze money laundering transactions without sharing data.
Money laundering anno 2024 is an art. Criminals who understand this art a bit use multiple transactions at different banks, often through other countries and foreign currencies, and sometimes with cryptocurrencies, to launder large sums of money.
All these tricks make it extra difficult to detect suspicious transactions. With multi-party computation (MPC), Rabobank (1) and ABN AMRO (2) have a new weapon to unmask sophisticated money launderers after all.
MPC offers banks the ability to use artificial intelligence to perform analysis on shared data, both nationally and internationally. Because the data is protected in a cryptographic way (for example, by multiplying it a number of times by a random number), no human or system gets to see that data.
"An ideal starting point, then, for uncovering money laundering through data analysis," said TNO, which developed the solution in collaboration with the Center for Mathematics & Computer Science and the two major banks.
Who gets to see the final outcome of the calculation is determined by the participating parties. A similar method is used by various government agencies in detecting tax fraud and other financial crimes.
(1) https://www.banken.nl/bank/rabobank
(2) https://www.banken.nl/bank/abn-amro
