By Glenn Smith, EMEA Head of AML & Head of Fraud, SAS UK & Ireland
Banks are being found out. As headlines continue to circulate regarding failures to meet tough regulatory standards, every other bank will be fearing the same predicament. With regulators determined to protect the interests of customers, accurately monitoring an ever-increasing volume of complex accounts is a daunting task for financial service providers. Yet, recent cases suggest that regulators feel anti-money laundering (AML) should already be well under control.
A systematic change is required within each bank not already capable of spotting and preventing suspicious transactions and customers among the millions of genuine ones. Only advanced and intelligent technologies can provide the accurate assistance needed to do this with ease. Ultimately, the only ones having a hard time should be those out to launder money!
A wave of false positives is setting banks back
Of course, banks already have AML solutions in place. However, research shows that the vast majority of alerts from these systems are false positives. Finding and dealing with the remaining genuine threats is a manual task requiring huge swathes of investigators’ time, sending costs skyrocketing and leaving the possibility that some bad actors will never be spotted.
Legacy rules-based transaction monitoring systems contribute to these inaccuracies. Though they are trained to spot the signs of money laundering, terrorist financing or sanctions violation, their simple, deterministic rules cast the net far too wide and clumsily for the ever-evolving nature of money laundering and fraud. What’s more, when new money laundering techniques for which there is no precedent emerge, these legacy systems are unable to adapt at pace before these practices start to spread.
When it comes to the recent cases in the news, while systems likely spotted accounts with unusual activity, one possibility is that they may have been lost among thousands of accounts and outdated rules. Far from proving the account holders to be criminals, this only proves that inaccurate AML technologies are all but redundant and unfit to satisfy regulatory demands.
Next-generation AML requires intelligent assistance
With 40 billion payments made in the UK in 2019, it’s clear that manual, inexact and time-consuming processes have no place in modern banking. The shift to pinpoint accuracy demanded by legislators would be impossible at this scale without advanced data management solutions founded on extensive AML expertise. In particular, technologies which use artificial intelligence (AI) and machine learning to identify and control risks are the key to finding only ‘true positives’ among the reams of transactions.
These intelligent technologies understand how money laundering and financial crime patterns look and evolve. Trained by experienced AML experts, over time they develop capabilities including data exploration and preparation, entity resolution, intelligent customer segmentation and natural language processing. They deploy these capabilities in banks as transparent scenario-based testing. Straight out-of-the-box, these scenario-based solutions can search every account for tell-tale signs of certain specific, modern money laundering techniques. Unsupervised, they flag problematic accounts to investigators who can then commit their time and resources to investigation of the criminals rather than the investigation of their own faulty systems.
Critically, AI and machine-learning-based systems must be accessible by all those involved on the ground with AML. They should be no-code, or low-code, enabling continuous training and updates by anyone regardless of IT capabilities. The intelligent systems should work and adapt in tandem with investigators to enable more meaningful data analysis – and therefore more effective AML – over time.
Transparent AML allows trust from all stakeholders
Many banks still fear the regulatory AML inspection. However, those with transparent, modern systems powered by AI have an understanding of how they’re effectively dealing with criminals. This knowledge creates confidence that banks require to protect their customers and keep the bad actors out, while demonstrating compliance with regulation.
Equally, the regulator must be confident in the bank. Once a suspicious transaction is identified and investigated, advanced AML technologies will allow the investigator to easily upload this alert and transform it into a Suspicious Activity Report (SAR). This report is then sent to the regulator electronically, with a full end-to-end audit trail from transaction through alert, case and SAR, before making contact with the reporting authority.
At the same time, it’s important that genuine customers aren’t impacted by ineffective anti-money laundering systems. The most intelligent AML solution will work in the background to keep the bank safe, without causing the friction to customers which arises from false alerts. With effective AML solutions, banks can conduct rigorous testing to spot risky transactions whilst ensuring that the average customer is barely aware of this powerful, intelligent system working behind the scenes.
Effective AML is a pain-free path to compliance
The key to effective AML is staying up to date with technologies that can support comprehensive AML techniques and capabilities. Legacy approaches leave competent teams at the mercy of historic rules, and team members may feel under pressure, overwhelmed by alerts, or powerless to improve a system which is making their working lives difficult. These attitudes are the result of being stuck with processes unsuited to the volume and pattern of transactions today, particularly as COVID continues to shape trends in financial crime. As the FCA indicates, legacy AML techniques should have disappeared long ago.
Adopting AI doesn’t mean teams have to abandon established ways of working. However, the extra analytical power brought to AML by intelligent technologies lifts the weight off teams and onto the vast experience and training embedded into these systems. By automating repetitive, manual processes, more accurately detecting suspicious activity, and putting these capabilities into the hands of those managing financial crime compliance, effective AML allows banks to serve genuine customers with confidence, and dishonest criminals with due prejudice.