On November 11, 2020, Joost van Houten (Managing Director of Sentinels, Slimmer AI’s financial risk and compliance business) was invited by Maurice Jongmans (CEO of Online Payment Platform) to present at VBIN’s annual compliance update. Well known within the Dutch payments industry, VBIN (United Payment Institutions Netherlands) is the leading trade association for payment service providers and electronic money institutions. Their members are a ‘who’s who’ of the innovative Dutch fintech community, including Worldline, Backaroo, IbanXS and Bitsafe and more.
As the Netherlands has one of the strictest compliance environments, the VBIN association serves several essential goals, including:
For the November 11thVBIN event, approximately 20 members of the association joined, a mix of risk and compliance experts and payments executives. The agenda focused on how financial institutions should respond to changes in the compliance landscape against the background of changing criminal behaviour, growing payment traffic and ever-increasing pressure from the regulator. The dynamic discussion and presentations ranged from regulations and industry trends, to human capital and technology.
The most fundamental task for financial institutions is to be compliant with the regulations, both today and well into the future. To achieve this, there are three questions that PSPs should ask themselves in order to be able to tackle this challenge:
For his presentation on AI in transaction monitoring, Joost van Houten explained how constant innovation and pragmatic use of the right technology can prepare compliance for in the future.
Joost outlined some of the key reasons why challenging the status quo in compliance is essential
The need to constantly innovate arises from the fact that criminals innovate themselves. Criminals learn on existing AML processes and develop even more sophisticated methods. In their September 2020 report, the Council of Europe MONEYVAL group has found that the covid-19 global pandemic made a surge in electronic transnational crime. As many are aware, more than $ 2 trillion is laundered worldwide every year and only 1% of it is detected. As criminals become increasingly innovative, so must financial institutions in order to combat them.
In many regions, particularly in the Netherlands, the regulator both expects and stimulates innovation. In 2017, DNB included in its Maturity Model (required to reach level 4 compliance in the Netherlands) requirements including automated pattern recognition and the use of network analysis to be part of the compliance setup of institutions. At the EU level, the European legislator is also busy. New regulations in the form of 6AMLD are planned for 2021, which also include specific asks around innovation and technological advancements.
for many innovative payment institutions, fast growth of their traffic has led to increased operational pressure. While it is a “good” problem to have, institutions are often forced to scale their transaction monitoring capabilities as they grow, and the tension in tech and compliance teams at smaller PSPs can become immense. In addition to the need for more suspicious transactions be properly detected and investigated, there is also the fear of ending up with a large compliance costs and high fines.
While on the surface, traditional solutions like "business-rules-only approach" have seemed compliant, this combined with the growing transaction volumes, have not only led to an explosive increase in false positives but also a low success rate in detecting money laundering cases.
Joost then presented the path Sentinels has taken in order to address the AML compliance challenges. With more than 10 years in applied-AI project and product development, Slimmer AI had a strong foundation in the technical capabilities required to build pragmatic AI solutions for this industry. Over the past 18 months, collaborations with Dutch leading PSPs, such as VBIN member - OPP, have further refined this approach in the AML and transaction monitoring solution - Sentinels.
Sentinels approaches the demands of transaction monitoring with the following four steps:
1. Data ingestion - integration of internal and external data sources.
2. Hybrid detection engine - a combination of business rules & machine learning.
3. Customer profiles - profiles of the modus operandi of customers based on peer groups.
4. User application - focusing on optimal usability and low or no need for training. Investigative alert handling and FIU reporting functions are optimised to be simple and seamless.
It is by looking at compliance and transaction monitoring holistically that AI can be applied safely and wisely.
During the VBIN event, members expressed interest in the use of AI in transaction monitoring and supporting their compliance efforts. But implementing this technology in such heavily regulated space can feel daunting. With their primary objective to be compliant to the regulations, PSPs that do not have significant expertise in artificial intelligence and machine learning are not sure where to begin with its safe and effective use. Senior leaders and risk and compliance experts also want to be able to fully understand and explain the decisions and recommendations reached.
Over the course of developing Sentinels, Joost expressed he’d heard similar concerns and how other payments leaders at other financial institutions are building personal and organisational confidence in applying new technologies and techniques. Software built with pragmatic AI supports and augments human abilities, with no intention to replace them. Pragmatic AI is only implemented for practical applications, which need human help to learn and adapt to business changes.
It is clear that constant innovation is required to help payment institutions stay compliant, become future-proof, and scale for accelerated growth. It is also clear that AI will be an important part of the future of risk and compliance. Increased knowledge and trusted partnerships seem essential for financial institutions to become more confident in applying AI and benefiting from the possibilities it can unlock. Putting humans at the centre of these important decisions will lead to better outcomes, including leveraging expert professionals effectively.
More about VBIN: https://www.vbin.nl/
To learn more about Sentinels go to: https://www.sentinels.ai/
To support your use of the website of Sentinels, we, Sentinels traded by Slimmer AI Sentinels B.V., need to register some of your information. We consider taking proper care of your data of utmost importance. We treat all information you provide as confidential.
In this privacy statement, we explain how we process your data. If you have any questions, remarks or complaints please contact us at email@example.com.
Your data is used for the following purposes:
Based on your data, the website of Sentinels does not make autonomous decisions that affect you or other people.
The website of Sentinels only stores your data as long as your account exists. You may delete your account at any time. If you use the website of Sentinels without a personal account, we only store the data you provide for as long as needed to be of service to you.
For security reasons, we store your IP address up to a maximum duration of 1 month. However, when a user attempts to access the website of Sentinels in a way that clearly differs from its intended use, this can be considered suspicious activity. As a result, the IP address may be stored for a maximum of 2 years for security purposes, and the IP address may be reported to organizations involved in internet security.
Any statistics based partly on your data will be stored, but cannot be used to identify you.
The website of Sentinels has no intention to process information of users younger than 16 years of age, unless they are permitted to use the website of Sentinels by their legal guardians. However, we have no reliable way to determine the actual age of users. We, therefore, advise parents to be involved in their children’s online internet activities to prevent any data from being gathered without their permission.
If you feel that we gathered personal data from minors without proper permission, please contact us at firstname.lastname@example.org. We will remove the data upon your request.
All your data is protected by state of the art technology. This ensures that your data cannot be accessed by parties who have no business doing so. If you feel that this is not happening in an appropriate way, please contact us at email@example.com.
Sentinels traded by Target Holding B.V. does not sell your data to third parties. Your data is only shared when this is needed for proper operation of the website of Sentinels, or to fulfil lawful obligations.
Any data we share with third parties is based on a data processing agreement with these parties. This agreement ensures that the parties process your data under the same terms that we do. Any shared data will be as restricted in scope and size as possible, given the purpose for which it is shared.
The website of Sentinels may also make use of third-party cookies like Google Analytics. These cookies allow us to analyze the operation and usage of the website of Sentinels.
You are entitled to view your data stored by the website of Sentinels, and you are entitled to have your data corrected or removed. You can send a request to do so to firstname.lastname@example.org.