AI Resilience Report forTellers

If you're worried about AI taking over teller jobs, the honest answer is: a lot of the routine work has already been automated — but not always by what people call "AI." For decades, ATMs, mobile apps, and check-counting machines have handled the cash counting, deposits, and balance-checking that used to fill a teller's day. Now AI is being layered on top of that infrastructure. The ICBA reports that AI is helping ATMs run more like smart, connected service points, with natural voice interactions, audio-guided experiences, and "AI-powered digital humans" that can handle staffing gaps and personalize the experience ^[1].

Vendors like Diebold Nixdorf and NCR Atleos use machine learning for predictive maintenance and fraud detection — for example, NCR Atleos says its Intelligent Diagnostics platform delivers repair guidance "with over 95% accuracy" ^[1]. At the human level, the role itself is shifting toward augmentation: community banks are reinventing teller positions into "universal bankers, relationship bankers or concierge bankers" ^[1] who use tech to focus on advice and relationships. Importantly, history shows technology doesn't always erase tellers — economists have noted the number of U.S. tellers actually grew from roughly 500,000 to nearly 600,000 as ATMs spread, before mobile banking began reducing branch visits ^[2].

Adoption is moving fast because banks see standing still as the bigger danger — a recent American Bankers Association survey found that bank leaders now view "doing nothing" on AI as the greatest risk they face ^[3], a major shift from earlier "wait and see" attitudes. Big banks are pouring money into it; Citizens, for instance, has described AI as speeding up everything from coding to customer service ^[4]. Costs are dropping as cloud-based AI tools become available off-the-shelf from vendors, making adoption realistic even for smaller community banks.

The labor market matters too: the U.S. Bureau of Labor Statistics projects teller employment will decline 13% from 2024 to 2034, with about 29,800 openings each year coming from workers transferring out or retiring ^[5]. But adoption isn't unlimited. Banking is heavily regulated, and the ICBA warns that community banks must carefully vet vendors on what data is collected, where it goes, and how the AI makes decisions ^[1].

Customers also still value human help for complex problems, fraud disputes, and big life moments like buying a home. The good news for young people: the skills AI can't easily replace — empathy, problem-solving, clear communication, and trust-building — are exactly the skills banks are now hiring tellers for. The job isn't disappearing overnight; it's evolving into something more relational and more skilled.