As artificial intelligence (AI) becomes more sophisticated and allows fraudsters to attack financial institutions in more complex ways, community banks may be wise to encourage their employees to look at all of their customer interactions in a more skeptical manner.
That was one of the takeaways during the FinXTech Session: Strengthening Your Bank’s Defenses, held earlier this month as part of Bank Director’s Acquire or Be Acquired Conference, in Phoenix, Arizona.
Matt Sickles, executive strategist at the IT solutions firm CDW, said the personal trust and service offered by smaller financial institutions could be exploited by bad actors who are using AI to impersonate the way bank customers write, talk and even look online.
“Any of you who have trained your own phone to be able to speak on your behalf knows how accurate it is,” Sickles said. “That will become one of the more threat-based events that will come in, especially into rural community banking — places we trust, people we know.”
Studies show the risk of fraud is growing due to AI. Microsoft’s Digital Defense Report 2025 showed that AI-automated phishing emails resulted in 54% click-through rates compared to 12% for traditional attempts.
Tara Schultz, senior vice president of strategic insights and industry relations at the core processor and technology company CSI, said things like AI-enabled voice cloning and video deepfakes are even more valuable to fraudsters.
“It’s designed to make us better, smarter, faster, more creative, more efficient — but also helps the fraudsters be better, smarter, faster, more creative, more efficient with your data, with account takeover,” Schultz said. “So, I would say as beautiful as that tool is, it’s also a risk that we all need to be aware of and look at how we’re bringing it into the good side of leveraging partners, leveraging solutions that can help us combat the bad side of that too.”
A New Type of Fraud
There is no question banks are concerned about fraud and the risks it presents. Cybersecurity was the no. 1 risk for respondents of Bank Director’s 2025 Risk Survey, and 69% said fraud was a top risk. More than 80% indicated they had provided ongoing staff education about the latest scams or threats and also implemented improved training on fraud.
But while 94% indicated their bank had fallen victim to check fraud over the previous 18 months, only 15% said the fraud involved artificial intelligence or deepfake media.
Steve Durney, vice president of partnerships and alliances for Quavo, a fintech company that assists financial institutions with fraud and dispute management, said during the session that more banks have certainly experienced AI fraud since the 2025 survey. “The days of getting the business email that has things misspelled in it and the wrong punctuation and the wrong addresses and things, that’s all over because they’re using AI tools as a means to do that,” Durney said. “And I think the community banks especially get vulnerable to those sorts of things.”
Just 21% who responded to the Risk Survey indicated their staff had received specific education on AI fraud and deepfake media fraud in the previous year.
Schultz said more banks are educating on this type of fraud now, and some are utilizing AI to prevent it, but many are not yet comfortable with that approach. “They don’t have the belief. They don’t have the trust,” she said. “It’s more of a wait and see.”
The LexisNexis® True Cost of Fraud™ Study 2025 North America backs up Schultz’s assertion. It found 44% of financial institutions rely mostly or entirely on manual fraud detection processes, while only 20% are mostly or fully automated.
The consensus among the experts on the FinXTech panel was that banks must fight AI fraud with AI to protect their customers.
“Think about the impact if you’re unable to gain access to your funds, gain access to your transactions. As a small-business owner, that can be one of the biggest risks that is facing our economy right now, if you can’t have that transactional access,” Sickles said. “Especially in community banking, where the security is less mature, we don’t have the available funds to start to close down on those risks. That right now is the omnipresent threat.”
Defending Your Bank
While funding limitations are certainly a concern for community banks in trying to defend against AI fraud, they aren’t insurmountable for institutions that have an organizational culture that values effective risk management. Durney and Schultz said banks that have buy-in and ownership of security protections from the C-suite down are best equipped to fight it. Finding the right technology partners to assist with that fight is also important. Sickles said banks no longer have to build fraud detection tools on their own, and should look at purchasing the right third-party solution that works for them.
Customer education is another key area where banks can help reduce AI-enabled fraud at the point of attack. Schultz gave the example of Williamstown Bank, a $228 million state bank based in Williamstown, West Virginia, which has an entire page on its website dedicated to security tips for its customers, and also consistently posts about fraud awareness on its social media channels.
“And it’s really important to hit all of your audiences, all of your ages,” Schultz said. “Their employees are really comfortable grabbing a customer and going to talk to them no matter what generation. So, I think that’s a great path that all banks should take.”
Banks that go the extra mile to learn from the scams they failed to prevent are also more likely to prevent the next one.
“I think the most critical thing is the consumer’s interaction with you at the earliest possible opportunity,” Durney said. “And then, secondarily, it’s harvesting as much possible data that you can from the event, learning about the consumer’s behaviors, learning about, ‘Did this come through an online channel? Did it come through a human channel? Did it come through social media? How did it originate?’ That allows your investigators and your prevention people to be armed with enough information to stop the flow and be able to react to it differently in the future.”
It also allows bank executives to figure out what technology solutions would prevent similar events in the future.
Even if community banks are potentially more vulnerable to AI fraud, when it comes to customer service resolving fraud, community banks may be at an advantage to bigger banks. “It is not just going to be a 1-800 number that the client calls,” Sickles said. “They’re going to want to go into a bank, sit down and resolve it.”
For more on fraud trends and security standards, Bank Director published The Fraud Menace: Protecting Your Bank in December 2025.
Previous