ork.AI
Forward-thinking banks, credit unions, and lenders know operational strategy must go beyond cutting costs, maintaining AML/BSA compliance, and other basic efficiencies. Global crisis after global crisis highlights just how critical it is for operations to also be scalable, flexible, and adaptable — in a word, resilient. Innovative approaches, like AI, help financial institutions combat money laundering and other financial crimes and lead to operational resilience.
Recently, Grant Vickers, head of Financial Crimes Strategy at WorkFusion, joined Nick Vitchev, Research Director at Chartis Research, the leading provider of research and analysis on the global market for risk technology to discuss the key role of AI-driven intelligent automation in banks’ operational resilience, particularly regarding CLM (client lifecycle management) and FinCrime. They also discussed how the use of these technologies can position an FI for new strategic opportunities that deliver revenue growth.
Here are some highlights from the conversation.
3 key compliance areas to target for AI, ML, and automation
There are three key areas of AML compliance operations that are benefitting greatly from the application of AI, ML and automation today:
- KYC (know your customer)
- Sanctions screening
- Transaction monitoring
KYC starts at the earliest stage of the client lifecycle – during customer onboarding. KYC informs an FI regarding who the customer is, why they opened an account, and what the FI is expecting of the customer in terms of normal activities. “A lot of activities will be based on the information collected and what FIs know about their customers so FIs can evaluate if they are operating as expected,” said Grant. “If their activities vary from the expected, then that’s a red flag.”
AI and Automation can help speed the customer onboarding and data collection process, but they can also surface good information about a customer – something analysts cannot do easily. As automation brings in information from third-party sources and open-source locations, AI and ML can delve into a wide range of documentary and non-documentary sources to extract key customer insights.
The combination of AI/ML and automation can also streamline the KYC process and simultaneously lower risk. “Most FIs are stuck in a periodic cycle where they look only at high-risk customers every year, then lower-risk customers every two or more years,” explained Grant. “Not only is that a huge operational burden, but you’re also going extended periods of time without reviewing most of your customer base.” So, banks are looking at tools to see KYC data in real time by scanning third-party sources for changes or receiving automated updates (review triggers) when unexpected activity occurs. “Banks want to move away from periodic reviews of KYC data, like once every 3 or 5 years, and instead, move to perpetual KYC, bringing information in real time and investigating only when a trigger event happens, like a change to a DBA name or a change in ownership,” said Grant. This type of AI-driven automation speeds the KYC process and also enables FIs to focus more on risk.
The second key area impacted by AI/ML and automation is more broadly understood – sanctions screening. “Screening customers against sanctions/PEP/adverse media are all ripe for intelligent automation, because banks can only tune down their screening so much, or else they’ll get dinged in the form of fines and penalties for letting too much through,” said Grant. “That leads to too many false positive alerts, lots of materially irrelevant news, etcetera.”
AI-driven automation can make decisions about false positives and the sentiment of news articles. Grant gave the example of how a screening tool would issue an alert for an article entitled “Bill Gates is Fighting Sex Trafficking in the UK”, classifying it as a negative article for review. But an AI/ML model would know that such an article is a positive mention of Bill Gates and automatically eliminate the need for an analyst to review it.
The third area impacted by AI/ML and automation is transaction monitoring for AML risk. Together, the technologies pull data from third-party sources to research organizations and individuals in support of determining if an alerted payment message (e.g., SWIFT message) is valid or a false positive. If a transaction is not determined to be a false positive (based on the AI model), the automation escalates to a human analyst for further review. AI and ML drive the conservative analysis applied to any alert investigation, then help to build narratives used to prepare for examinations for regulatory filing.
Alleviating Worker Shortages – Level 1 Analysts
Ever since the start of the Covid pandemic, banks have been faced with high attrition rates in compliance operations, noted Nick. Some banks have even had to double hiring over the last 18-month period, putting stress on training, retention and maintaining the organizational structure. That in itself offers a key driver toward intelligent automation, especially regarding the need to handle spikes in L1 alert volume. This situation has only continued to worsen, as the Russian invasion of Ukraine has led to thousands of new sanctions being rolled out by the United States Treasury and other governmental bodies around the world.
”We’re seeing lots of attrition among analysts, but banks are also having a really hard time hiring analysts. L1 analysts are getting burned out on mundane work with lots of information, yet rarely finding escalation-worthy items. So banks struggle to find talent,” stated Grant.
Bank compliance operations leaders see that analysts doing higher level work are happier in their jobs. “So they ask us how to enable them to focus only on meaningful work, allowing them to work on risk-based decisions which are more valuable to the analysts and the bank. That’s why we designed Digital Workers like this, to help them focus on those high-value things,” explained Grant. “When you bring in a digital workforce, you help solve those problems, because they never leave, they work 24×7, and they get really good at the tasks you give them over time, thanks to ML.” That solves issues related to operations capacity, quality, attrition, and customer experience (CX). He explained that CX improves, because intelligence built into the Digital Workers reduces the number of customer touchpoints needed to disposition an alert.
How automation helps day-to-day AML operations beyond crisis events
Nick noted that, beyond one-off crisis events, “there are many more mundane issues that are forcing organizations to think long-term and think toward automation.” As an example, he said that he recently spoke with a European bank that closed a relatively small acquisition that barely affected their top line. Yet, the relatively small volume increase in L1 alerts – combined with the need to re-screen their customer portfolio – made their FinCrime Operations “fall apart.”
Grant understood the situation, noting that many banks state today that their examiners are wondering why they haven’t applied automation in the 3 areas of KYC, sanctions screening, and transaction monitoring. “It’s ripe for intelligent automation. It’s not event-driven, but examiners are looking at AML CLC programs and saying they see an opportunity. The script used to be flipped with examiners having to be convinced, but now they are recommending automation,” noted Grant. This gives banks much more incentive to pay attention to intelligent automation.
As for bank M&A, the issue will grow as banks merge, their geographic coverage and customer base expand, and there’s an exponential increase in the amount of compliance work that results. “The KYC of the other bank’s customers remediate if the KYC isn’t up to standard, and new tech to handle the increases. Automation plays a key part in this,” said Grant. He also noted that many FIs are hesitant to bring in automation during M&A transactions, instead of waiting until after the merger is completed. “But we say to consider it as you go through the M&A process. As processes change or you think about a new operating model, it would be helpful to bring your automation vendor and other technology vendors along to have a single cohesive process,” he explained. An area of value occurs when bringing your customers and customer data to another platform. That typically introduces many data integration issues. “So that’s ripe for automation, whether an exposed API or going into the front-end of an app,” Grant noted.
Turning KYC knowledge into strategic revenue opportunities
Some banks and other FIs see increased benefits from a strategic perspective when they leverage intelligent automation. “Not just efficiency…more streamlined customer experience, quicker payments, faster decisions. You can even repurpose that rich profile data and behavior to make strategic decisions. Do you see that?” asked Nick.
Grant agreed entirely. “We bring in automation to understand and classify the documents, gather information insights, and digitize the documents to make them ‘alive’ versus stagnant. Now they have millions of discoverable documents that are great for compliance. But also, revenue centers can offer new products and services based on what they know about their customers.”
Grant further explained that Many FIs are moving to RTP (real-time payments). “But when a sanctions screening process hangs up a payment and puts it on hold until a compliance review happens, banks need automation. With wires you could take a day, but RTP has to be seconds,” he added.
As a final example, Grant noted how WorkFusion’s automated KYC capabilities have led COOs at top US banks to openly discuss their customers being frustrated after already handing over many KYC documents – only to have an analyst from the other side of the world ask them for unnecessary information. “So you bring in technology that enables the analyst to get the information without having to ask the customer. Those are the 3 big buckets of solving revenue generation problems on top of regulatory compliance problems.”
A recent report from WorkFusion and Chartis Research presents findings from interviews with Tier 1 and 2 banks and FIs on the key role of automation in their operational resilience, especially when it comes to CLM and FinCrime. Click here to download your copy.
Schedule a demo of our AI Digital Workers.