The obligations that banks face to combat financial crime, both from regulation and internal policy alignment to strategic values, is placing increasing cost burdens on banks. Nordea, as a GSIFI and with over 10.7 million customers, is no exception. As criminals employ increasingly sophisticated techniques in laundering the proceeds of their crimes to evade detection, so banks like Nordea must continually increase the sophistication of the tools and techniques they use to meet their obligations.
Today’s mainstream technology used to detect money laundering is based on rules designed around techniques criminals are known to use. It’s then up to human investigators to look more closely into each case and determine whether or not there may be foul play. Typically, in over 95% of cases flagged, investigators conclude there’s not. Again, Nordea is no exception.
This is extremely wasteful, expensive and inefficient. It’s also demoralizing for investigators, leading to high staff turnover and compromising their ability to be truly effective.
So, our long-time client Nordea asked us if we could help. Their question, how can artificial intelligence help us reduce artificial threats?
Bringing together teams with expertise in transaction monitoring, regulatory and advanced analytics from Sweden, the United Kingdom and Switzerland, we produced a tailored approach to deliver a proof-of-concept for Nordea.
Using our Swiss-based innovation labs, leveraging leading edge machine learning capabilities, we worked with Nordea’s experts to design a powerful model to detect and remove this waste, whilst also highlighting activity that needs immediate attention.
This model integrated neatly with Nordea’s existing systems and was taught to recognise what Nordea’s investigators had previously considered to be both high and low risk alerts.
Like the algorithms that help online retailers target customers according to their taste in brands, the model we built with Nordea became more insightful and accurate as it learned how to interpret the signals that indicate criminal activity, analysing vastly greater quantities of data, more reliably, than existing people or processes could.
In fact, it was able to say with a high degree of certainty and without compromising the detection rate of criminal activities, that around 40% of the cases highlighted were not worthy of immediate investigation. This has a very significant impact on human investigators, reducing their workload and enabling them to focus their attention where it counts.
These results are in line with our client’s expectations, but show potential to deliver substantial additional business benefits, reducing costs, increasing efficiency and effectiveness, whilst helping the bank to stay on the right side of changing regulations.
The financial services industry is undergoing rapid transformation creating new opportunities but also threats. At EY we are passionate about the use of AI and machine learning to generate growth and reduce risk.
For financial crime specifically, AI is protecting our clients and their customers, but also by reducing time spent on artificial threats, it’s enabling investigators to use their time more efficiently and enriching their jobs.
While there’s much discussion about AI in financial services, we recognize that many companies are at the start of their journey. That’s why we’ve made significant investments not just in technology but in bringing the right people and teams together to help answer our clients’ biggest questions.
We train and nurture these inclusive teams to develop minds that can transform, shape and innovate financial services. Not just for today but the future. It’s how minds made for financial services, contribute to building a better working world.