Dilok - stock.adobe.com
Danske Bank fights money laundering with AI
Danske Bank is using artificial intelligence to help it prevent a repeat of the huge money laundering scandal that hit one of its operations in Estonia
Danske Bank is continuing its investments in anti-money laundering (AML) technology with an investment in artificial intelligence (AI) technology to monitor and investigate potential money laundering.
The Nordic bank is now using AI-based technology from data and analytics company Quantexa, which it began piloting in 2018.
The bank has been on a mission to transform its anti-money laundering capabilities after substantial illegal transaction activity was uncovered at a Danske Bank subsidiary in Estonia.
The Baltic state made global headlines when Danske Bank was found to have transferred more than €200bn of suspicious money through its tiny branch in Tallinn.
Danske Bank’s transaction-focused AML-detection apparatus proved sluggish in identifying account and money transfer divergence anomalies at the branch in Estonia’s capital.
During the pilot of the platform from Quantexa, Danske Bank improved its ability to detect suspicious activity in its market trading business in areas such as foreign exchange, securities and equities.
“Harnessing technology enables us to identify complex financial crime behaviours more effectively. Running advanced analytics on a wide range of datasets can help us better detect, investigate and prevent financial crime,” said Satnam Lehal, head of financial crime detection at Danske Bank.
The bank has now taken the pilot project to the next stage and integrated it with existing infrastructure.
Read more about the battle against money laundering
- Danske Bank improves its anti-money laundering software, utilising artificial intelligence and machine learning.
- The top 5 benefits of AI in banking and finance.
- Money laundering was back at the top of the agenda recently when the EU’s Fourth Anti-Money Laundering Directive came into force.
Danske Bank is not alone in addressing anti-money laundering shortfalls. In April 2020, Swedbank announced it was overhauling its AML IT infrastructure in the wake of a highly critical report that identified deficient AML detection, control and customer validation systems as the key factors that contributed to the Swedish bank’s systematic failure to prevent money laundering-related transactions in its Baltic subsidiaries.
The report, conducted for Swedbank by London-based law firm Clifford Chance, found that €36.7bn in transactions, all carrying a high risk for money laundering, were processed through the bank’s branch network in Estonia, Latvia and Lithuania between 2014 and 2019. In March this year, regulators in Sweden and Estonia imposed fines totalling €347m on Swedbank for breaching money laundering laws.
According to the UN, about $2tn is moved illegally each year. Money laundering was recently headline news when leaked documents revealed that banks including HSBC, Barclays and Standard Chartered have moved huge amounts of money despite spotting suspicious transactions.
Criminals use big banks to hide their dirty money, which is often linked to organised crime, with funds being used to pay for assets to hide the money’s origin.
In the Netherlands, ING was fined €775m in 2018 after the regulator said the bank had failed to prevent the laundering of hundreds of millions of euros between 2010 and 2016.
Meanwhile in 2017, Citigroup agreed to pay almost $100m and admitted to criminal violations as it settled an investigation into breaches of anti-money laundering rules involving money transfers between the US and Mexico. In the same year, Deutsche Bank was fined $650m by British and US authorities for allowing wealthy clients to move $10bn out of Russia.
Under the threat of huge financial penalties, banks have turned to technology to detect money laundering activity. Today, machine learning and natural language processing are being used to replace manual work. Machines can read many more articles than humans and can automate anti-money laundering processes.