Community Bank Voices: AI and the Human Value Proposition

Community banks are built on personal connections. While technology is core to service, the human touch remains a top draw at smaller institutions for customers and staff alike.

Preserving this people-centric sensibility is vital, say community bank leaders, as artificial intelligence begins infusing operations and performing tasks people once did. Though usually following their big-bank counterparts in AI uptake and implementation, community banks increasingly are embracing a future with AI and are laying the emotional and practical foundations for its use.

“We are starting with a philosophy about how we’re implementing AI so that our employees and teammates can get comfortable that the goals for AI projects are not job elimination,” said Dawn Mugford, chief risk officer at Norway Savings Bank.

It’s no wonder, these leaders say, that their workforce is worried about a technology often described as deeply human. Generative AI, after all, can replicate human communication with speed, accuracy, and scale. Agentic AI is learning to do what people do, increasingly without prompting. The temptation is to use this powerful capability broadly in the interest of cost savings, and overhaul banks top to bottom.

But that’s not the plan at his bank, says David Stewart, chief credit officer at Kleberg Bank. “One perspective is to create a culture where you’re using it to reduce headcount. Another, and this is where we’re planning to take it, is to increase the efficiency of existing staff,” he said.

A Culture of Usage

“Efficiency” can mean different things to different people. For some, it’s synonymous with cost-cutting. For others, it’s making it easier and quicker to perform everyday tasks. Given already tight staffing numbers, heightened sensitivity to customer and employee experience, and the strong desire to protect their value proposition, community bank leaders seem to agree that efficiency, in the case of using AI, means empowering instead of replacing people. 

Few, meanwhile, see ignoring AI as an option. In ProSight’s 2025 Community Bank Survey, for example, 80% of respondents said that using AI effectively would be critical to meeting strategic objectives over the next five years. But they are also playing catch-up. In ProSight’s 2026 CRO Outlook Survey, 68% of respondents at banks with less than $50 billion in assets said they’d not developed AI upskilling or training programs for their workforce (vs. 24% for larger banks). Almost half said they’d not built AI technology infrastructure either.

But that’s quickly changing, Stewart believes. “There might be some leaders who aren’t forward-looking and want to keep doing things the way they always have. But if you’re sitting in this [executive] seat, it’s hard not to see what’s around the corner,” he said.

Familiarizing the organization with potential applications of AI, getting executives and employees comfortable with the technology, and painting an energizing picture of how employees can work with AI are all part of the setup work happening at some community banks. “AI is a great tool for generating conversations about how we do things as a bank,” Mugford said. “When you’re a smaller institution, there’s a lot you do manually, but as you start to grow you have to consider more automation. Where it makes sense, you’d much rather have people doing analysis [and other high-value functions].”

In this way, AI has potential as a talent-development tool. “We love it when people stay a long time. AI can create opportunities for folks to train and do different things; to be able to do some projects we maybe haven’t had capacity for before,” she said.

Companies use “human in the loop” to describe an arrangement in which people oversee AI outputs and apply higher-level judgments to create a final product. The term has also become HR shorthand for “no job losses.” Creating a culture where people are users and beneficiaries, and not victims, of AI supports community banks’ ambitions to adopt it, Stewart suggested.

Including people is practical and risk-focused, too. Half of CRO Outlook Survey respondents said that using AI without adequate human verification would be a top AI-related risk for their organization.

Vendors and Customers

This vein of concern runs through employee, vendor, and customer approaches. Community banks run on limited technology budgets and depend on vendors for off-the-shelf products and capabilities. In the case of information-intensive AI, marrying internal stores of data with external large language models presents thorny data custody and privacy issues banks must manage.

“We’re really raising our game in vendor due diligence to make sure we’re asking all the right questions about how our data is being shared, stored, and used in model training,” Mugford said. When a vendor’s practices don’t readily align with the bank’s guiding principles on data usage and storage, it’s a sure sign to reconsider that partnership, she added.

The right vendor products are shortcuts for small banks to new AI capabilities. Use in customer service workflows and fraud detection frameworks is already happening at bigger banks, which may have the resources to build their own tools. Community banks depend more heavily on vendors and are extra-selective about their use cases, the bankers suggested. Still, they see areas such as loan origination and monitoring as fruitful targets for the technology—always with human validation embedded in the process.

When it comes to customers’ awareness of and experience with AI in the banking context, the less apparent it is the better. Consumers want seamless services that deliver to their expectations. If AI performs a customer-facing task poorly, it can destroy a bank’s service reputation almost instantly. If it replaces a valued human interaction, it can undermine the personal-touch value proposition.

“A big part of how we differentiate is knowing your customer and having a personal interaction,” Mugford said. Theoretically, AI could support these personal connections: “If you had, for example, detailed informational prompts at your teller lineup such as birthdays or other data…they could help you create or deepen those relationships with your customers,” she said.

Banks big and small are wrestling with similar issues as they reframe growth, operations, and strategy through an AI lens. For their customers, AI is invisible infrastructure; it’s the outcomes in service that matter. AI adoption at community banks must fit around relationship banking, not redefine it.

“When vendors are properly vetted and managed, the customers don’t care if it’s AI that’s helping to protect them from fraud or delivering a better product. They want to be able to pick up the phone and talk to somebody and have intelligent touchpoints like a great mobile experience. That’s how we differentiate,” Stewart said.