Banks don’t know their customers anymore due to digital transformation, which has serious implications when fighting financial crime. Refine Intelligence CEO, Uri Rivner explains how they can now use automated outreach and AI to regain this superpower.
The 80s! What a fantastic decade. Just consider the movies: Indiana Jones, Karate Kid, and the iconic Back to the Future.
Let’s go back in time together. It’s 1985, and you’re walking into your local bank branch. As you walk in, your friendly banker immediately recognises you. Hey Joe! she says. How’s the family? Still, planning that trip to Europe? I hear Katy is going to a top college - would you guys need a loan for her tuition?
Banks have existed for centuries, providing financial services to consumers and corporations. During most of this time span, interactions were face-to-face, allowing a degree of familiarity with customer needs that amounted to a real superpower: Really Knowing Your Customer and their constantly evolving lives.
Let's move forward in time, into the current world of digital transformation. Hey, don’t get me wrong – digital transformation is awesome, enormously convenient, super quick, and extremely user-friendly. It also saves a huge amount of operational costs as branches close, replaced by digital services.
But there’s an often-overlooked side-effect: the new relationship model is one-sided. The customer punches some keys in the app, telling the bank what to do - but not the context behind it. Losing the two-sided interaction means a sharp decline in a bank’s ability to know the customer.
How bad is it? Here’s a question for you:
When the Anti-Money Laundering (AML) Investigation Officer looks at an alert and finds no explanation, they may ask the line of business to explain the activity. When such an inquiry goes to the customer’s branch, what are the odds the branch can answer it without reaching out to the customer?
The answer is 12%*.
In 88% of inquiries, the branch chases the customer for answers. The branch doesn’t know the customers the way they used to – and if the branch doesn’t know, who does?
AML Transaction Monitoring systems certainly don’t have that sort of context, so it’s no secret that the vast majority of anomalies that trigger AML alerts are perfectly legit customer activities. As a result, the daily routine of the AML Investigation team can be quite frustrating: looking at false positive after false positive.
Can banks restore their old superpower, and gain the much-needed knowledge of their customers’ life events so they can quickly resolve alerts generated by legitimate activities?
One way to find out what’s behind an anomaly is by asking the customer. Today this is done by chasing customers over the phone; this is extremely ineffective, especially for consumers and small businesses. Instead, banks move to an automated outreach program in which customers answer such inquiries using digital channels. Here are some best practices:
Choose digital communication channels your customers prefer.
Engage in UX analysis and a/b testing to optimise the digital outreach experience.
Select the right questions to ask, optimising between explainability (how much context they provide) vs. user experience (what % of customers answer the questions);
Use a structured set of questions, rather than free-form answers. This will help you build benchmarks and analyse typical customer behaviours.
Use the right tone - too friendly might trigger not taking the inquiry seriously, while too firm might cause the customer to stop responding.
Consider when to ask for documentation, using a risk-based approach.
A more advanced approach is using Machine Learning to automatically recognise anomalies that are actually genuine customer activities and provide sufficient evidence to ‘green-flag’ them. This requires tapping into a labelled data set of confirmed genuine - that is to say, falsely flagged anomalies and their legitimate explanations.
Why is the customer moving a lot of funds into the account and doing a big international wire transfer? A good AI model will tell you that the most common explanations are perfectly legit real estate or investment activities. Why is the customer pulling thousands of dollars in cash from their account all of a sudden? The top reasons are related to buying a vehicle or doing a renovation project, where contractors ask to be paid in cash.
Like in any AI system, the devil is in the details - but the general approach is modelling the good, not just the bad.
To summarise: whether by using automated inquiries delivered via digital channels, or modelling life events that can explain anomalies, banks are beginning to restore the long-lost superpower of knowing the customer.
* Based on a Top 50 US Bank using Refine Intelligence to automate compliance-related customer outreach
About Uri Rivner
Uri Rivner is the CEO and Co-founder of Refine Intelligence, a fintech startup that is disrupting the paradigm of fighting financial crime with its new paradigm of ‘catching the good guys.’ Refine Intelligence uses AI to automatically identify the real-life story behind anti-money laundering alerts, providing full explainability and reducing alert handling time by 90%. Uri was co-founder of behavioural biometrics company BioCatch and Head of New Technologies at RSA.
About Refine Intelligence
Refine Intelligence helps banks move to a new paradigm of fighting financial crime by ‘catching the good guys.’ Our mission is to enable banks to regain their ‘superpower’ of understanding customers’ life events that create changes in their financial activity. The company was founded by financial crime-fighting experts who have built innovative game-changing companies in online fraud detection.
Every day we send out a free e-mail with the most important headlines of the last 24 hours.
Subscribe now