How AI will disrupt, enhance ATMs

AI can benefit ATMs in three key ways: availability, security and hyper personalized experiences.

Photo: Adobe Stock

February 28, 2025 by Bradley Cooper — Editor, ATM Marketplace & Food Truck Operator

AI is transforming industry through generative language models, operations, security and much more. The ATM industry is also feeling its impact, but it also has a key opportunity to maximize the benefits of AI.

Stuart Mackinnon, COO at NCR Atleos discussed how AI will disrupt and transform the ATM industry at a session at the ATMIA conference held in Orlando from Feb. 5 to 7.

Mackinnon said when we think about AI’s use cases in ATMs, we think of three main categories: availability and service, security and hyper personalized experiences.

For repairs, Mackinnon said that ATM manufacturers have tried to “convince you for decades” that they have predictive analytics to fix ATMs. But he emphasizes, “Predictive is almost impossible. No one wants to fix a part before it breaks.”

However with AI, predictability with repairs becomes more of a possibility.

“If I see this happen and this happen, 9 times out of 10 this will happen. I can be more clever about my repair,” he said.

This can also take the form of AI dispatch and call routing for service workers. In many cases, ATM manufacturers have to hope they have sent the right service worker for the issue the ATM is having. With AI, it can “correlate all of the data and get the right person there.”

Once the service worker is there, AI models can provide information for the repair or audit the completion of the repair tasks to ensure it was done properly. It can use computer vision to do this via an app.

AI can also help with ATM security. While it’s not going to stop a car from ramming into it, it can help with identifying key behaviors via existing security systems.

“We can use cameras much more effectively when we can detect malicious behavior before it happens. If I can detect the difference between a car backing up to it vs. a car that drives past it. I can proactively send alarms, call people, dispense intelligent banknote neutralization system,” Mackinnon said.

It can also detect certain hand signals such as if someone is trying to insert a skimming device, or detect when someone is operating an ATM under duress, such as a customer is being held at gunpoint and told to withdraw money.

“Stopping stuff after the fact, we have been trying that for decades. Improving things during or before the act is what we are focused on,” Mackinnon said.

When examining hyper personalized experiences, Mackinnon said he doesn’t believe AI is going to be a landscape change in the ATM business. He said that every four years there’s a major shakeup in the industry such as advertising looping screens or bitcoin, but neither ultimately makes a huge difference, but they do lead to incremental improvement.

He believes AI will be similar in its ability to develop hyper personalized experiences. For example, AI can improve voice operated ATMs via Azure OpenAI GPT-4o, which can understand people’s accents, languages and dialects and answer questions naturally. This also allows customers to authenticate at the ATM with their voice.

In the long term, some ATMs may be able to deliver customized customer engagement both at the bank or at a retail ATM. For example, a retailer could integrate their customer data into the ATM and be able to offer customers customized offers, such as shampoo coupons or advertisements on the display.

Mackinnon emphasized we are only “scratching the surface” of what we can do on the ATM. However, it is key that ATM operators utilize AI and innovative tools in general, as banks continue to push bank services away from the branch to outside services such as ATMs.

Table of Contents

About Bradley Cooper

Bradley Cooper is the editor of ATM Marketplace and Food Truck Operator. He was previously the editor of Digital Signage Today. His background is in information technology, advertising, and writing.

Connect with Bradley: