Some small regional banks in the United States are stepping back from the race to adopt AI. They worry about customer data leaks and cybersecurity risks. They also judge that AI could weaken face-to-face service, a core competitive strength for local lenders.

American Banker reported on Nov. 10 that U.S. community banks are choosing a strategy of adopting AI in stages after verifying security and customer convenience, rather than refusing to adopt it outright.

That contrasts with large banks that are increasing AI investment. Bank of America said in July it had approved more than 300 uses of AI and machine learning for investors. Some regional banks, by contrast, are prioritising staff training and setting up internal control systems.

Blake Chatelain (블레이크 샤틀랭), chief executive of Red River Bank, stressed that while he recognises the usefulness of AI, there is no need to be a first mover. He said the bank would move quickly once safety, security and customer benefits are proven.

Red River Bank, with assets of $3.3 billion, is interested in using AI for fraud detection and back-office automation. For now, it is focused on strengthening cybersecurity against AI-based hacking and financial fraud and on setting up an approval process for employees' use of AI at work.

It is also building a system to check what information employees enter into AI services. The measures reflect concern not only about external attacks but also about the risk of customer data leaks caused by employees' carelessness.

Other regional banks are also taking a cautious stance. Bank Five Nine in Wisconsin is testing AI in a limited way, and First United Bank in Texas is training all employees on the risks of AI and the potential for misuse.

Chesapeake Bank in Virginia said it worries that replacing customer service work with AI could cost it its strength as a people-centred service. Texas Bank and Trust in Texas also takes the view that technological innovation should not replace relationships with customers.

Chatelain said customers may gather information using AI, but for major financial decisions such as buying a home, taking out a loan or planning retirement, they ultimately want to consult directly with a banker they trust.