Skip to main content

Anti-money laundering requires an automated approach

A tango between humans and technology

Governments are trying to combat money laundering through laws and regulations. To comply with these requirements and avoid fines and reputational damage, financial institutions have hired large numbers of colleagues. According to the Dutch Association of Banks, this amounts to around 13,000 bank employees, or 15 percent of all those employed in financial services. Yet, even with this workforce, it is still insufficient to adequately monitor the enormous volume of transactions and customer data.

The implementation of these laws and regulations seems to be becoming a societal issue. Shouldn't these employees be better utilized in education or healthcare? In my opinion, the answer is yes. The approach to combating money laundering can and should be more effective, with fewer people but better trained.

Not more, but smarter

Money laundering practices do not always occur in the same way. Criminals are not only technically savvy, but they are also well-informed about the protocols and screening mechanisms used within banks. Staying one step ahead of them requires experience and domain knowledge, such as in the areas of Know Your Customer (KYC), money laundering, and fraud. This requires specialists who can improve processes by leveraging data and technology in a smart way. In my view, the solution lies more in increasing the impact of experienced specialists rather than hiring more and more employees.

Clearing the path to automation

In October 2022, the administrative court ruled that online bank bunq is allowed to use artificial intelligence and data models to monitor customer transactions. The Dutch Central Bank also advocates for a more risk-based approach and increased efficiency through the use of technology in its report "From Recovery to Balance," published last year. The ruling and the report create an opportunity for financial institutions to automate controls to a greater extent. This has fortunately paved the way for the use of smart technology, optimally supported by data. However, this does not mean that humans are becoming obsolete. On the contrary, the combination of systems and humans is essential. Tech and touch.

The human added value

In setting up an effective process, humans continue to play a crucial role. The various risks must be weighed and translated into the design of systems and the use of software. What notifications do the systems provide? How reliable are the predictive signals, and what insights can be gained from this information to take action? This involves not only investigating unusual transactions but also constantly refining processes.

When ICT systems are fed with the right characteristics by domain experts, they can significantly increase effectiveness. By quickly aggregating and comparing large amounts of data, and by recognizing patterns that are deeply hidden in the data and not visible to humans.

The human touch is important in assessing such patterns. An analyst is able to interpret an anomalous pattern: is it related to criminal activities or a crowdfunding campaign? It will continue to be a dance between humans and technology.

Future perspective

This is not yet a piece of cake. There is a risk of bias, a human prejudice that can creep into the software. It is therefore wise to have additional eyes, preferably from outside the organization, oversee the management of the systems. Experts can assess what adjustments are needed in data models for the desired risk assessment. For this, it is important that we continue to learn. Only then can the use of technologies such as machine learning and artificial intelligence help to manage the risks of money laundering with a smaller number of specialists.

Author: Mark Breed, Practice lead FEC Ordina

Take also a look at our whitepaper on financial economic crime

Let's talk

This field is required

This field is required

Please fill in a valid email.

This field is required

This field is required

This field is required

This field is required

By submitting your message, you consent to Ordina processing the information in accordance with our Privacy & Cookie policy. We take your privacy very seriously.