The Fraud Arms Race: Why Automation Alone Won’t Protect Merchants

Why Isn’t Automation Enough to Stop Fraud?

Fraud has always been part of doing business, but throughout 2025 it has become faster, more targeted and far harder to contain. Criminals are no longer relying on crude tactics or volume alone; they are using automation, artificial intelligence, and stolen data to imitate genuine customers with alarming accuracy. At the same time, merchants are under ever-increasing pressure to reduce friction, approve more transactions and protect the customer experience, all while keeping fraud and chargebacks under control.

Automation has become the default response to fraud. Across e-commerce and digital payments, machine-learning models, automated decisions and real-time transaction scoring are widely used. However, fraud losses and chargebacks remain stubbornly high, partly driven by the continued rise of friendly fraud and first-party misuse across the UK.

This is what defines the term – ‘the fraud arms race’. As technology improves on both sides, merchants are discovering a difficult truth: while automation is essential, it is not enough on its own. In fact, the businesses that really stay ahead are those that balance smart systems with human oversight, judgement and experience.

Why Are UK Merchants Facing More Sophisticated Fraud Than Ever?

Fraud in the UK has changed shape. Traditional card fraud is still a problem, but criminals are now taking advantage of gaps created by rapid digital growth, online sign-ups and real-time payments. With a strong e-commerce market and high card usage, the UK remains one of the most targeted countries in Europe.

A lot of this change is driven by how easy fraud has become to access. Through Fraud-as-a-Service, criminals can buy ready-made tools that automate attacks, test stolen card details, switch identities and copy normal customer behaviour. As a result, fraudulent activity can become almost indistinguishable from genuine shopping.

At the same time, new security rules like Strong Customer Authentication (SCA) haven’t removed fraud so much as shifted it. While some types of unauthorised card fraud have fallen, more disputes now come from genuine customers claiming transactions they actually made. For merchants, these cases can be especially hard to spot and challenge.

Combined with the growth of subscriptions, digital services and instant delivery, fraud risk is now built into everyday customer activity. It’s no longer just about stopping obvious criminals, but about managing behaviour that looks legitimate on the surface.

If Automation Is So Powerful, Why Isn’t It Solving the Problem?

Automation plays a vital role in modern fraud prevention. It allows merchants to screen thousands of transactions per second, apply consistent decision-making and respond instantly to known risk signals. Without it, most online businesses simply could not operate at scale…However, automation has blind spots.

Automated systems learn from historical data. This means they are very good at spotting patterns that look like previous fraud, but are less effective when fraud evolves or deliberately mimics legitimate behaviour. It’s all about keeping up with the criminals at large, and the fraudsters who understand this constantly adjust their methods to stay just below automated thresholds.

There is also the issue of false positives. In the UK market, where competition is high and customer loyalty is fragile, incorrectly declining legitimate customers can be just as damaging as letting fraud through. Automation often errs on the side of caution, which can leave genuine shoppers frustrated without merchants fully realising why.

Perhaps most importantly, automated tools cannot truly understand intent. They can identify anomalies, but they cannot judge context. Was a transaction disputed because it was fraudulent, because the customer forgot about a subscription, or because a family member used the card? These distinctions matter greatly in chargeback management, but even the most advanced algorithms do not have the capability to detect them.

What Risks Do Merchants Face When Relying Too Heavily on Automation?

When automation becomes the sole decision-maker, risk doesn’t disappear – it simply changes.

One major risk is misclassification. Friendly fraud disputes, which now account for a significant proportion of UK chargebacks, often pass through automated filters undetected until the dispute stage. At that point, the damage has already been done: fees have been applied, ratios impacted and valuable time lost.

There is also a compliance risk. Card scheme rules and dispute requirements change regularly, and automated systems aren’t always able to keep up or handle the grey areas well. When transactions don’t follow familiar patterns, human review is often what makes sure the right evidence is submitted and that it meets the latest expectations.

Another common blind spot is context. Automation is good at showing what is happening, but it doesn’t always explain why. Without people reviewing the data, merchants can miss early warning signs such as new abuse patterns or changes in customer behaviour that could be prevented with clearer communication, better billing descriptions or improved customer service.

Over time, this can create a false sense of security. Everything looks like it’s under control on the surface, while losses slowly build in areas that automated systems aren’t designed to catch.

Why Is Human Oversight Still Essential in Fraud and Chargeback Management?

Human judgement allows for perspective – something technology is yet to achieve.

Experienced fraud and chargeback teams understand nuance. They can review edge cases, assess customer histories and recognise when a dispute is more about confusion or dissatisfaction than criminal intent. This is especially important in the UK, where consumer protection expectations are high and disputes are often raised as a first point of contact.

Human oversight also plays a crucial role in feedback loops. Decisions made by analysts help refine automated models, making sure systems learn from real-world outcomes rather than assumptions. This partnership between people and technology is what keeps fraud strategies relevant as tactics evolve.

In chargeback management specifically, human expertise is often the difference between recovery and write-off. Knowing which cases are worth fighting, how to frame evidence and when to concede strategically requires experience – not just data. This is one reason many merchants quietly benefit from specialist support, even if it operates quietly behind the scenes.

It’s important to note that human involvement doesn’t mean slowing things down. When used correctly, automation handles volume and prioritisation, while people focus on judgement-heavy decisions where their impact is greatest.

How Can UK Merchants Strike the Right Balance Going Forward?

The most resilient fraud strategies are not built upon choosing between automation and human input, but rather on integrating both effectively.

Automation should act as the first line of defence, handling tasks like transaction processing and anomaly detection. Human teams can then focus on oversight, managing exceptions and driving strategic improvements. This layered approach enables merchants to scale efficiently without losing control.

For many businesses, striking this balance goes beyond internal teams. As fraud and chargeback rules become more complex, some merchants choose to partner with specialists who focus exclusively on dispute management. This isn’t about outsourcing responsibility, but about supplementing in-house expertise during peak volumes or when handling particularly complex cases.

Just as importantly, fraud prevention should never operate in isolation. Insights from disputes can improve customer communication, clarify billing, and even influence the design of products. It could be said then, that when humans interpret this data, those insights are far more likely to lead to meaningful, practical change.

As fraud can never be eliminated in its entirety, the goal isn’t perfection –  it’s resilience. Merchants who understand where automation ends and human judgement begins are better positioned to adapt as the arms race continues.

So, Will Humans Still Matter in the Fraud Arms Race?

The fraud arms race shows no signs of slowing. Automation will continue to improve, and fraudsters will continue to exploit it. In this environment, relying on technology alone is not a strategy – it’s a risk.

For UK merchants, the future of fraud and chargeback management lies in collaboration: systems that are fast and scalable, guided by people who understand context, regulation and customer behaviour. Automation can process transactions, but it cannot protect trust on its own.

In conclusion, it is the combination of automation and human oversight that provides stability. While fraud prevention is often framed as an arms race, it is not one that either side can win alone. Automation delivers speed, scale, and data, but human judgement turns those tools into decisions and strategies. Therefore, in a world where fraud is increasingly automated, it is this interdependence, not competition, that keeps merchants one step ahead.