Banks Turn to AI to Stop Overstocking ATMs
Cash has not disappeared the way many predicted. Cash accounts for 14% of consumer payments in the United States, more than 80% of consumers used it in the past 30 days, and 90% expect to keep using it, the Federal Reserve said Aug. 4 in its 2026 Diary of Consumer Payment Choice. That persistence leaves […] The post Banks Turn to AI to Stop Overstocking ATMs appeared first on PYMNTS.com .
Despite initial predictions, cash remains a key player in consumer transactions, accounting for 14% of payments in the U.S. and used by 80% of consumers in the last month, according to the Federal Reserve's 2026 Diary of Consumer Payment Choice. This persistent demand creates a dilemma for banks: overstock ATMs with cash and tie up capital or risk running out, triggering extra costs or customer frustration.
Artificial intelligence is beginning to solve this problem by turning it into a forecasting issue rather than a guessing game. Companies like H2O.ai use AI to create cash-demand models for ATMs, analyzing historical withdrawal patterns, paydays, holidays, and seasonal trends to forecast cash needs with an accuracy of about 15%. This precision allows banks to stock ATMs more accurately, reducing the need for excessive buffers and freeing up cash that can earn returns.
Brink’s takes this a step further by integrating cash supply, branch inventory forecasting, order optimization, and cash monitoring into a single system. By predicting cash needs seven to ten days in advance, Brink’s can reduce total cash demand across its ATM portfolio by 30% to 40%. This is achieved by anticipating spikes or dips in demand, allowing for better inventory management and reduced transportation costs.
AI is also being applied directly to the ATMs themselves. Hyosung Americas uses AI to analyze historical transaction data, nearby events, and seasonal fluctuations to determine the optimal cash balance for each individual ATM in real time. This minimizes idle cash while reducing the frequency of cash refills. Additionally, Hyosung uses AI to monitor machines for early signs of mechanical issues, enabling technicians to diagnose and fix problems with a single visit instead of two.
The adoption of AI extends beyond the ATM. According to the PYMNTS Intelligence report "Time to Cash™: A New Measure of Business Resilience," 70% of surveyed firms use AI for cash flow management. Those using agentic AI have automated up to 95% of their accounts receivable processes, compared to 38% for businesses without AI integration.
This widespread adoption of AI in cash management underscores its potential to revolutionize how banks handle ATMs, reducing costs, improving efficiency, and enhancing customer satisfaction.
Written by urgent.news from PYMNTS's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
