Latest bank risk report points to AI use as increasing cybersecurity, fraud concerns

Advanced “superintelligence” (SI, but known more widely as artificial intelligence (AI)), as employed by “threat actors” amplifies cybersecurity and fraud concerns and is thus a key risk to national banks, their regulator said Thursday.

The Office of the Comptroller of the Currency (OCC) said in its Semiannual Risk Perspective (SARP) for fall, 2026, that other key risks to national banks include:

  • Higher interest rates could affect both the net interest margin (NIM) and the valuation of securities held by banks. The report also suggests higher rates could also increase commercial refinancing risk.
  • Credit risk in the loan portfolios remains modest overall.

Regarding threats related to AI, the report noted that the “growing availability” of sophisticated AI models “has dramatically changed the fraud landscape.” It said AI “can further enable threat actors to carry out various types of fraud such as social engineering, account takeover, or impersonation with greater speed, scale, and complexity.”

“A comprehensive approach to fraud risk management, as well as the ability to respond to emerging threats of fraud can help deter fraud affecting a bank and its customers,” the report stated.

It recommended Internal controls designed to prevent and deter AI-enabled fraud utilizing a layered approach. “Examples of approaches include incorporating human review and using a different communication method from the one used in the request to help verify a request is legitimate (e.g., ‘out-of-band confirmation,’ or video verification). In addition, monitoring the speed or volume of an activity such as a login attempt, password reset, or transaction can detect if unusual activity is happening too quickly or too often (e.g., behavioral and environmental analytics, velocity monitoring).”

However, the report also suggested that AI can be leveraged by “banks of all sizes.”

“Standard security and anti-fraud strategies, including access controls and identity verification, can help mitigate” threats by bad actors, the report suggested. It added that AI can also support banks’ efforts to address cyber threats and fraud, using AI capabilities to process large amounts of data and recognize subtle patterns.

OCC’s Semiannual Risk Perspective Highlights Key Risks in Federal Banking System

Be the first to comment

Leave a Reply

Your email address will not be published.