Despite the initial pushback against AI from compliance departments, it now seems the technology allows for being both fully compliant – and efficient.
The numbers show that banks are bearing significant costs to keep up with AML policies. According to a Bloomberg report from 2018, FIs allocate 5 to 10% of their revenue to compliance, which makes for an average of $20 billion spent in a year.
To make matters worse, global regulations keep changing. The EU anti-money laundering directives come one after another: the 5MLD has been in effect since January; still, the 6MLD is already on the table.
1 trillion saved
There’s hope though: by employing AI, we can bring the said cost down in the long run. As stated by Autonomous Research, in the next 12 years, applying AI to KYC, AML workflow and other areas of data processing, can result in savings of more than 20% . Overall, AI is projected to save the banking industry approximately $1 trillion by 2030. It’s not surprising then that already 35% of organizations, as shown in the Digital Banking Report, have deployed a machine learning solution supporting at least one of the following processes: AML transaction monitoring, fraud identification, sanction screening and KYC checks.
To fully benefit from the savings, financial institutions have to make sure that the technology they wish to employ is compatible with their operations. It must be decided to what extent AI is actually needed, and possible to implement, taking the existing IT and compliance structures into account.
AI comes with many benefits: it can identify patterns which are far too complex to be detected by the human eye or rule-based monitoring. Employing intelligent solutions in transaction monitoring and detection modeling decreases the cost of manual labor, time spent on monitoring, and increases the result accuracy. In this context, combining artificial intelligence with traditional AML system can boost the efficiency and accuracy of the risk detection process.
Transparency first
What is more, the great advantage of using AI results from its ability to accelerate AML compliance. One of the main regulatory doubts has been raised around the transparency of AI working principles. Explaining why a certain case is marked as a suspicious one is very challenging but crucial in order to meet compliancy standards. At this point, applying the so called Explainable Artificial Intelligence provides results that are possible to interpret and explain. In this context, Explainable AI is the key to meet compliance requirements for more accurate and productive models.
Another field in which AI helps to meet compliance standards is the use of unsupervised learning for anomaly detection. Finding the inconsistencies within analyzed data at the right moment may save banks from severe financial penalties for omitting money laundering cases. Some more complex and unusual behavioral patterns are hard to spot or completely undetectable to compliance employees. Since intelligent algorithms are able to identify suspicious patterns quickly, financial institutions have the chance to stop criminal activities before they escalate. This is basically the way in which the old and common problem can be solved by the newest technology.
Regulators say ‘yes’
Regulators in the United States already gave the green light to emerging technologies and innovative applications for risk management in 2018. Specifically, Joint Statement to Encourage Innovative Industry Approaches to BSA/AML Compliance has been issued. The statement recommends that banks should “responsibly” implement and use AI-based approach to meet AML requirements. The regulators consider new technologies helpful in reporting money laundering and terrorist financing. They also agree that artificial intelligence and machine learning could provide greater banking strategies to manage money-laundering and terrorist-financing risks more effectively, reducing the cost of compliance at the same time.
On the European level, the Nordic-Baltic financial sector declared its support for the development of an extensive AI plan. Their EU Policy Recommendations for years 2019-2024 underline the importance of conducting business in a trustful way for both customers and society. To achieve this goal, they aim at developing a well-functioning and reliable financial system through regulations promoting innovation and digitalization to let banking businesses grow. Specifically in the AML context, the Estonian Banking Association reports the necessity to use modern information technology means to monitor transactions in order to establish more effective real-time screening and post-transaction monitoring.
Challenge accepted
With the rapid development of new technology, fast growing money-laundering schemes and more people making cross-border transactions, it has become a challenge to keep traditional AML systems effective. Since money laundering methods become diverse and keep evolving, it is a high priority for financial institutions to maintain a great level of confidence in AML and risk management. We believe the way to achieve this is the use of AI – as it brings measurable cost benefits and the opportunity to meet compliance standards much better.
Learn more about Comarch Anti-Money Laundering solution
Account Manager in Finance, Banking and Insurance Sector at Comarch S.A.
