Chris Oliver from Lysis Group stated that the global prevention of money laundering remains a challenge for both regulators and firms. This is due to criminals becoming more sophisticated with their money laundering tactics and firms having to deal with large volumes of complex customer data. In fact, according to the UK Treasury Select Committee, the amount of money being laundered through the UK, or its structures, is estimated to be tens of billions of pounds or even hundreds of billions.

This calls for a more targeted approach in KYC practises which can be achieved by introducing financial crime technology, specifically artificial intelligence (AI) and machine learning. AI can be defined as technology which can create intelligent systems that mimic human intelligence and behaviours, and machine learning can be described as a subset of AI which allows machines to learn from data, using algorithms, without specific programming. According to Chris, “AI can add enormous value in the anti-money laundering space because it can analyse large amounts of data much faster and more accurately.  AI can also identify possible trends or patterns of money laundering that the human eye might miss”.

According to insights from the International Monetary Fund, a wider adoption of AI in regulatory technology (RegTech), is due to firms becoming more competitive in the financial space. They stated that AI can effectively reduce false positives which normally creates the bulk of AML/CTF alerts. This will enable firms to have a more dedicated focus on cases that are in fact suspicious. Technology has come a long way Historically, the majority of AML interventions required a lot of detailed human involvement. Along with this, the open web was utilised to screen individuals and transactions, but the output produced volumes of unnecessary noise, which required analysts to conduct intensified manual screenings to identify possible suspicious individuals and transactions. The risk linked to this approach resulted in many suspicious transactions not being detected and various flagged individuals going about their business undetected. Therefore, over time, money laundering and terrorist activities have increased dramatically which forced global regulators to become more stringent in their regulatory approach. This domino effect resulted in firms searching for more efficient and effective ways to track money laundering activities with the objective to remain compliant. Also, with the growing number of people that are gaining access to the global financial system, cross-border transactions and diversified payment systems are making it harder to keep track of illicit transactions and related activities, which is where AI and ML can play a significant role.

The benefits linked to AI, as part of the screening process, are multiple because the element of human error is reduced.  AI also enables analysts to insert more specific information regarding individuals or entities which increases the level of accuracy of information dramatically based on the related risk parameters. Slow adoption of technology “One must realise that many firms have legacy systems due to technology changing so rapidly over time”, Chris pointed out. This tends to complicate the implementation of new AML technology because compatibility remains a general challenge. This could result in ‘silo’ operations within firms where systems and processes are not in sync. Furthermore, as with any new technology, AI also poses some limitations which are linked to risks.

According to the Financial Action Task Force (FATF), new technology should be used responsibly and manual reviews, along with human input, remains a winning combination. A key strategic decision Chris concluded by saying that “Although there are many AI technology vendors in the market, there are only a few that offer the right capabilities needed in the AML space. Also, AI technology varies quite a lot, with each having their own strengths and weaknesses, which leaves no room for a one-size-fits-all approach. The Lysis Group can assist firms in selecting the right AI technology which is aligned with their bespoke operational nuances because we know first-hand that, over time, firms develop both generic and unique operating characteristics. Therefore, as a key strategic decision, it is vital to partner with a technology vendor that offers the right ‘fit’ over the long-term.