4 AI use cases transforming the future cash cycle

Whether you live in an advanced or an emerging economy, cash remains a critical pillar of public infrastructure. This is true even as digital payments grow worldwide. It is also true that we live in a dynamic time. The payment landscape is no different. Cash has also undergone yet another seismic transformation with the advent of the digital revolution, especially with the use of artificial intelligence. AI has found its way into the discourse around the cash cycle, with its proponents extolling its operational and cost-saving benefits.

While these benefits are no doubt true, AI is at a stage now where it can help central banks in evidence-based decision-making, using insights gleaned from across the cash cycle. This would impact every part of what they do, from demand forecasting to in-vault operations to the logistics of getting cash to actual users. In effect, AI can act as a planning and orchestration layer, enabling central bankers to transition from a reactive position where they are “dealing” with situations to a creative and strategic posture, with pragmatic planning in place for every contingency.

Let us see how such an approach might work in practice, with four real scenarios.

Forecasting

Printing and issuing currency are clearly central to what a central bank does. Overprinting raises problems of storage and ties up capital, while not having enough money in circulation is an issue as well. (No central banker wants to see long lines of worried and impatient citizens outside banks and ATMs.)

While statistical models and expert judgment drove forecasting earlier, recent shocks have exposed shortcomings. Sharp spikes in cash demand – or even a surge in demand for a particular denomination – unsettle markets and user sentiment.

Forecasting tools that incorporate AI can offset this. Their strength and flexibility allow them to model for, among other things:

• Signals such as denomination demand
• Seasonal spikes and troughs
• Payment behavior that is driven by structural drivers and unforeseeable one-off events, such as a natural catastrophe 

With AI, production and issuance can be planned across denominations while factoring in numerous stress scenarios. Machine‑learning models can be trained to generate multiple outlooks, including multi-year views that offer central bankers a bouquet of options to choose from, using their own expert judgment. Leadership gets the data and visibility it requires to plan effectively.

This is in addition to the wealth of historical data a central bank would already hold, and its deep understanding of macroeconomic variables in their own ecosystems. Resilience across the cash cycle is built up because contingency plans are in place before the pressure appears.

Reduced overprinting, enhanced denomination availability, and better budgeting are just some of the positives of such an approach. By using AI responsibly, central banks can pivot to a dynamic production strategy that is resilient and flexible over many years, while adhering to the principle of human judgment and control.

Local visibility and inventory management

Among other challenges, central bankers are faced with these deceptively simple questions: where should banknotes be, in what denominations, and when? Printing notes is just the beginning of the process – they must reach the user as well.

While central banks forecast nationally, demand can be highly localized, especially in larger nations. Demographic markers play a role: demand could peak around a particular community’s religious observances, for instance. Central banks can err on the side of caution when budgeting for these, and with good reason. But there is clearly potential to reduce inventories and costs and model demand more effectively, without sacrificing resilience.

AI can help with this, by accounting for indicators such as withdrawals, lead times, and lodgments at the level of the individual vault and branch. A bank would carry a leaner inventory and order fewer “emergency” transfers. Shortages and surpluses can be anticipated before they occur, and stock can be balanced proactively, without affecting resilience in any way.

Operations, logistics, and workforce management

AI can have a transformative effect within the cash center and its related activities as well, whether centrally or at a more localized level. Banknote management, including sorting and destruction, is highly mechanized, but the planning of workloads and staff is often based on rough historical averages. Daily volumes can be volatile and are impacted by cash-in-transit (CIT) company schedules, regional events, or even shifts in retail patterns that cause changes in demand.

Among other things AI-driven solutions can model for:

• Historical volumes of incoming and outgoing notes
• The routes and schedules of commercial partners such as CITs
• The effects of foreseeable events such as paydays and public holidays
• Increasingly, operational data from connected machines in the cash cycle, including throughput and downtime

The result is more accurate data across sets that include cash centers and vendors, and also markers such as denomination demand and even note quality. This opens up improvements across staff scheduling, better allocation of machines, and now, predictive maintenance of machines around times where lower demand is expected. Further, skilled resources (which tend to be scarcer) can be allocated where they’ll actually be required and not distributed across the network.

An interesting development that AI and increasingly powerful computer networks have raised is anomaly detection. “AI can analyze vast datasets to detect unusual patterns quickly, flag anomalies for detection, and support decision-making,” states a recent Bank for International Settlements (BIS) report.¹ In effect, AI provides an additional layer that scans for subtle anomalies which might otherwise escape manual review. This provides further signals that central bankers can read to “predict” the future.

Think of the cash cycle, and how it was traditionally conceived. A cash center might have been seen as primarily a processing hub. But with AI, it becomes – along with every other data-generating component of the cycle – a sort of sensor network, where intelligence about the ecosystem is continuously being collected, analyzed, and fed back to the central bank.

Close the loop: from operational data to policy

This is perhaps the most powerful contribution AI has to offer central bankers in charge of the cash cycle: it lets them close feedback loops between operations and strategy.

High‑frequency data from the cash cycle – including order volatility, delays, imbalances, return flows, ATM outages, and the like – allied with recurring stress patterns can carry early or “faint” signals of deeper shifts. AI helps humans to find and read these signals, and to place them in their proper contexts.

Modern data platforms and AI-driven solutions can integrate real‑time operational data with macroeconomic indicators. They can design and run simulations, for instance the effect of a reduction in ATM networks on user confidence, or the impact of a one-off event, such as the Iberian blackout of spring 2025, on cash availability.² They can provide dynamic dashboards that decision‑makers depend upon for readily accessible data across various metrics.

When using AI in this way, issuance planning can better reflect what is happening on the ground. Positive outcomes can range from better print timing to tighter and more data-informed stock levels. It can offer decision-makers visibility across the entire cycle. In so doing, blind spots are removed, which means fewer “shocks to the system.” AI can help strategy to become more evidence-based, and thus transparent.

Governance remains key, of course. The responsible use of AI in financial services can only be built upon a robust framework anchored in human oversight and best-in-class risk management, so the public is best served and trust is maintained across the cash cycle.

Using AI responsibly to deliver a more resilient and effective cash cycle is thus an imperative for central bankers. Cutting through the noise around AI is equally important to delivering the best possible outcomes for a nation’s cash cycle and its citizens. 

G+D understands the requirements of this situation. It has been part of over 150+ cash cycles around the world, from printing and issuance to logistics, cash center operations, and destruction of old notes. It understands the dynamic trends that are shaping both the cash cycle and the larger payment landscape. As a security technology pioneer, it is at the forefront of developing cutting-edge solutions that use all available tools, including AI, delivering them to where they can be best utilized across fields such as payments, connectivity, and identity. 

Published: 11/08/2026