TNO, Rabobank and ABN AMRO work on privacy-friendly data analysis

News article
Article tags:
  • Innovation
  • Detecting Financial Crime

Sharing and analysing data is a fundamental part of detecting financial crime. The more relevant information we can analyse, the more effectively our crime analysts can work. While our clients’ privacy is paramount, an innovative approach could reconcile these two key goals. The Dutch scientific research organisation TNO is testing this together with Rabobank and ABN AMRO. A new system using ‘fake data’ is showing promising results.

New technologies have huge potential in the fight against financial crime. But it requires cooperation between several parties, as many past initiatives have shown. Also, these technologies are still in the fledgling stages of development. To find out more and to assess whether they could prove useful as tools for detecting financial crime, ABN AMRO andRabobank are collaborating with TNO on a project called Multi Party Computation for Anti Money Laundering (MPC4AML).

Good initial results

The project has produced good initial results based on a synthetic data set – data that has the same characteristics as regular data but doesn’t relate to real clients. The synthetic data set contains fictional clients who make payments between various banks. The clients in thisdata set have a risk score, like real clients. Clients whose accounts are linked to salary payments and mortgages usually have a lower risk score than, for example, those who receive large cash deposits.

We don’t want to share these risk scores with other banks. But if one of our low-risk clients receives money from high-risk clients at other banks, we want to be able to monitor that client more closely. So how can you do that if you’re not allowed to see the risk scores used by other banks? The technology that we are testing in this collaborative project could hold the answer. By encrypting and splitting the data, we can ensure that nobody can find out the original risk score.

Nevertheless, using an algorithm, this system can calculate which of our low-risk clients are involved in transactions with high-risk clients at other banks, and vice versa. This is important information that can help us deploy our analysts more effectively in future. That way, lesstime could be spent on clients that analysts check under the current system, and more could be spent on others.

Combating money laundering and terrorist financing

Experts estimate that around 2,400 billion euros worth of criminal transactions are carried out worldwide each year. Often these relate to money-laundering of profits from drug dealing and other criminal activities. But countless other illegal activities such as terrorism, human trafficking and illegal wildlife trade are also financed by criminal money. Banks have a gatekeeper role in the financial system. This means that we have a legal obligation to identifyunusual transactions and, by doing so, contribute to a safer society. That’s why we actively combat money laundering and terrorist financing.