Banking sector, regulators announce joint effort to address AI risks

Banking industry and Treasury Department representatives came together today at the American Bankers Association’s headquarters in Washington, D.C., to announce the next steps in a public-private collaboration to combat artificial intelligence-related fraud and cybersecurity threats to the financial system.

The Financial Services Sector Coordinating Council in March released a list of recommendations for the industry on AI as part of a larger Treasury report on managing AI cybersecurity risks in the financial sector. That collaboration will continue, with monthly meetings where participants will discuss topics such as ways the financial sector can leverage AI and close the capability gaps between large and small institutions.

“Cyber criminals and fraudsters are not new threats of course, but with the advent of artificial intelligence, the challenges they pose will only increase as the Treasury’s recent report on managing AI-specific cyber and fraud risks laid out,” ABA President and CEO Rob Nichols (pictured, left) said. “The good news is that banks are not new to AI. Our sector has been using AI and machine learning for longer than most industries. We know that bank investments in technology and innovation can help us fight back against the bad guys to protect our customers and the broader financial system.”

AI technology is evolving at a rapid pace, said FSSCC Chair Deborah Guild (pictured, second from left), head of enterprise technology and security at PNC Financial Services Group. “With this speed comes the probability of increased risk,” she said. “So not only are we challenged by the mounting threat related to AI advancements, but we must also be attuned to the potential vulnerabilities this technology could represent as we use it to combat this threat.”

The Treasury report lists 10 next steps, said Todd Conklin (pictured, second from right), chief AI officer at the department. They include addressing the capabilities gap between small, midsize and large institutions, the need to narrow the fraud data divide, and expanding the National Institute of Standards and Technology AI risk management framework. “In the past, we cooperated and were able to collectively identify challenges, but we did not materially address these challenges together in a regulatory manner. That is no longer the case,” he said. “We now have a model for how to implement positive solutions for our toughest challenges, from cloud to AI and beyond.”