Convergence Of Financial Institutions’ Risk Management Aided By Artificial Intelligence
The attention on technology and risk in financial institutions has often been laser focused on transactions. That makes sense because of the large volume of trades and other transactions that are at risk from fraud, hacking and other issues. However, financial institutions also have other risks. Organizations have been slowly increasing the use of artificial intelligence (AI) in other areas, and there are signs that AI is supporting a convergence of risk management tools in financial institutions.
Financial transactions, trading in particular, have long been a focus of technology. The faster institutions can trade, the more competitive advantage they have and the higher profits. That’s meant significant money spent on colocation, networking and other technologies. It has also meant that automated trading was one of the first area in finance where AI has been applied. That then led directly to AI being leveraged for fraud detection and prevention, as the rapid analysis and flagging of transactions was required.
What has been happening in the last year or two is that risk managers at financial organizations are looking at multiple aspects of both risk and AI in order to understand how to create a better, more holistic, view of risk. “Mobile devices have massively changed how customers interact with financial organizations,” says Matt Tengwall, General Manager of Verint Fraud & Security Solutions. “Yet even with those, physical structures and other touch points still matter. Risk managers must look widely at how technology in general, and artificial intelligence, in particular, can help lower risk.”
Tools That Lower Physical Visits While Managing Risk
While mobile payment numbers have grown much faster outside of the USA than within, there is still growth. In addition, financial institutions have been strengthening their apps and slowly attracting more mobile customers. That adds a new layer of risk, in both networking and device security.
Customer call centers, in all industries, are being improved with AI. Sentiment analysis is improving the ability of call center personnel, and chatbots, to appropriately respond to customers. A recent expansion of sentiment analysis has been past pure text to analyzing tone, meter, and other verbal cues to improve the understanding of a customer’s mood. One of the key risk management enhancements is understanding how written and verbal cues can help call centers identify potentially fraudulent calls. An AI system can be trained to grade calls based on potential risk and for escalation to a backline team for a more focused handling.
Those Visits Still Happen And Risk Management Is Changing
While those are semi-obvious changes in how the financial institutions are changing, one of the more intriguing areas discussed with Mr. Tengwell is the area too many people overlook: the physical banking infrastructure. While the number of people showing up to banks is steadily decreasing, it is not going away soon. While upper-middle class and wealthy pundits like to talk about the disappearance of cash, working Americans know differently. Cash matters. There are also still requirement to talk to bankers about loans and other specific services that require in-person meetings.
Yes, the physical infrastructure is changing, with fewer few service offices, the expansion of limited service branches in grocery stores continues. So does the need for ATMs and even some kiosks providing services in between ATMs and branches. All of those require security.
One area where technology and electronic money have helped is in branch security. While cash is needed, less is required in banks. Regular camera-driven security is being enhanced with AI to quickly identify threats. More importantly, regular risk management procedures mean that you don’t interrupt a potentially violent person. It’s better to let the person leave the branch and then deal with it. Combining AI recognition of potential risks by using visual and verbal information can limit personal risk. That technology can also potentially assist or replace the teller alert switch, again improving overall safety.
At the automated level, linking what’s been learned about fraud through analysis of major trading systems can be pushed closer to the customer, with in-person banking, ATMs, and mobile devices.
“People have recognized the need for AI at the transactional level,” said Mr. Tengwall. “What’s now being seen is that the convergence of physical security and other risk management tools and processes is an important issue.”