Only 1% of laundered cash in EU is detected — ABN AMRO wants to improve that
ABN AMRO’s Ivich Hoffman reckons detecting money laundering is like mixing the perfect cocktail
ABN AMRO takes a holistic approach to detecting laundered money
Analyzing financial data with AI, advanced analytics, and machine learning technologies plays a major role in that task, sure, but so do manual investigations and other forms of cooperation between banks.
“It’s not only about looking at information (transactions), it’s also about putting the right questions to customers,” said ABN AMRO’s managing director of financial crime detection Robin de Jongh at TNW2020.
“So, being curious, but also holding a neutral way of thinking towards the customer, and not assuming that everyone is a crook or criminal,” he added.
To stay ahead, ABN AMRO runs its ownDetecting Financial CrimeAcademy to help investigators sharpen their skills.
De Jongh: “We employed 25Sherlock Holmes-typesto detect financial crime via intelligence-led analysis on money laundering schemes and modus operandi.”
Tech won’t stop, so staying ahead of the curve is crucial
On the other hand, ABN AMRO’s Ivich Hoffman likened her bank’s method of detecting laundered cash to making the perfect cocktail.Hoffman,Product Owner Connect for Detecting Financial Crime,was one of the panelists during TNW2020.
“In my opinion, there are three main ingredients: data, advanced analytics, and cooperation between banks,” said Hoffman, before adding that one of ABN AMRO’s slogans is “data is at the heart of everything.”
The three ingredients affect more change as the sum of their parts. Data, for one, needs to be as broad as possible. “If we only use our data for one bank, does the crime stop at the door of that bank? Or is it better to see data of all banks together?” asked Hoffman.
(We talked about this in more detail in the podcast ‘How banks detect money laundering.’You can listen to it here.)
There’s also the need for targeted approaches when analyzing potentially suspicious data (this is the advanced analytics part). For example, if one investigates all “Politically Exposed Persons” for potential involvement in money laundering, many suspicions will be proven correct. But this isn’t always the case.
“Therefore, it’s much more effective to create models,” said Hoffman. “For example, within ABN AMRO, we have used a data science model to detect potential human trafficking, and we are using artificial intelligence and machine learning models to detect unknown unknowns.”
Banks are gate-keepers that must work together
In some sense, the third ingredient is the most crucial: co-operation between banks, and that’s exactly where TMNL comes in.
The way ABN AMRO sees it, banks are “gatekeepers of the financial system,” and so those institutions have a statutory duty to protect the integrity of that system.
Criminals and their financial networks, however, are considerably complex. Those handling illicit funds often bank with a number of banks simultaneously, and may bounce between them in a bid to hide their dodgy dealings—making it even more critical for banks to share their algorithmic firepower.
“Do you think it’s effective for banks to develop all these models on their own? Or is it better to cooperate?” asked ABN AMRO’s Hoffman at TNW 2020. Personally, she’s leaning towards the latter. “In my opinion, the best way to detect financial crime is in cooperation with other banks. And that’s what we have started.”
Story byDavid Canellis
David is a tech journalist who loves old-school adventure games, techno and the Beastie Boys. He’s currently on the finance beat.David is a tech journalist who loves old-school adventure games, techno and the Beastie Boys. He’s currently on the finance beat.
Get the TNW newsletter
Get the most important tech news in your inbox each week.
