AI is materially transforming how financial institutions understand and manage their own businesses. It has also been game-changing in terms of the products they offer to their customers, and the safeguards they put in place to protect against fraud and money laundering.
Financial criminals are increasingly sophisticated, intelligent and many times better-funded than individual organisations in the financial services industry; they’re constantly probing for ways to capture, disperse and obscure the movement of illicit funds though payments networks. “We’ve observed funds being split up and moved around so they look legitimate, deliberately designed to avoid detection by traditional rules-based systems,” said David Rich, Executive Vice President of Services at Mastercard. “Collaboration on our live, network level solution has enabled the participating banks to act on insights and share these learnings amongst the group, and also to identify other insights or deliverables they need to help them,” he added.
Other patterns that have been detected by Mastercard’s financial crime solutions to trace illicit funds include cycling between accounts to boost their credit rating with the holding bank, and the incredibly rapid movement of funds, of separate transfers from a single account in two minutes. “We don't believe humans are fast enough to do this,” noted Rich. It demonstrates the behaviours of an AI driven automated script.
“AI is imperative to success, or the contagion of financial crime will continue to expand.”
— David Rich, Mastercard
Developing solutions that are capable of combatting these practices is particularly critical in a world where money moves in real-time. “We’re forever trying to catch up,” said one attendee. “Before you had time to check a payment: today you have no time at all.”
Account-to-account payments are the perfect vehicle for applying machine learning technology to aid in the prevention of financial crime. And unlike manual processes, it’s near-instant: AI can be used to verify such things as account name to reduce authorised push payment fraud or accidental misdirection of payments, or whether an individual or business is attempting to send funds into a suspect mule account in a split-second — before the payment is even initiated. “AI is imperative to success, or the contagion of financial crime will continue to expand,” Rich asserted.
And what about the advent of Open Banking, which introduces new players, including non-traditional third-party providers into what has previously been a relatively closed relationship between a financial institution and its customers? There’s concern that third-parties won’t have invested in cyber security, AML and fraud prevention in the same way that banks have over the years, introducing new risks and a certain amount of unpredictability around how these new participants are likely to behave. “We have the technology, it’s easy to use if you have the skills, and the data is essentially the same,” offered Salla Franzen, Chief Data Scientist at SEB. These same solutions can be used to secure access to consumer data in the wider digital ecosystem.
From financial crime, the discussion moved onto how else banks can employ AI and data science to model scenarios that enhance their business operations, such as liquidity forecasting.
“My job is to inspire the bank to be data-driven.”
— Salla Franzén, SEB
“The biggest risk [financial institutions] run around the issue of liquidity is that of reputational risk: of hitting your cap and ending up on the news,” said another attendee. “In an unstable economy, if a debit cap is breached and there’s no money in the ATMs, there are riots.”Financial institutions need to be able to understand not just what has happened and why quicker than ever before, but also to project what might happen.As the world moves into real-time gross settlement, every facet of banking will be impacted by the ability to forecast accurately liquidity.
A huge increase in the volume and standardisation of data creates an opportunity to automate what were previously manual processes. The huge volumes of data at banks’ disposal improves the accuracy and effectiveness of these solutions. The adoption of AI tools and techniques, including machine learning, robotic process automation and natural language processing, can radically transform the way financial institutions are structured and function, if banks can capitalize. “My job is to inspire the bank to be data-driven”, explained Franzén.
