How Banks and Credit Unions Can Get Ahead of Synthetic Identities and AI-Driven Fraud

Fraudsters are embracing AI to deploy fraud attacks faster than many financial institutions can respond, setting the stage for a 153% surge in fraud losses between 2025 and 2030. Juniper Research estimates the value of those losses will reach $53.8 billion, up from $23 billion last year.

In this Q&A, Chris Harrison, CEO of Fideo Intelligence, explains how fighting AI-powered fraudsters more effectively not only demands more robust data signals but also the ability to act on them in nanoseconds.

How can banks and credit unions keep up with fraudsters who use AI to perpetrate fraud than many organizations are keeping up with today? That was the central question during a panel discussion at Fintech Meetup 2026. This Q&A with Fideo Intelligence CEO Chris Harrison is adapted from that conversation.

Q: Fraud losses are climbing fast. What has changed in the fraud landscape for banks and credit unions?

A: “The biggest shift is that fraud is now being driven by highly organized, well-funded operations rather than isolated bad actors. These are sophisticated groups—cartels, state-backed organizations, and networks involved in terrorist financing—that run fraud like a business function. They have clear objectives, well-defined KPIs, and specialized teams dedicated to tasks like data acquisition, social engineering, and monetization.

They are also embracing automation and AI to move faster. Instead of a single phishing campaign or one-off account takeover, they run continuous, data-driven workflows. They test, optimize, and scale attacks across institutions, which means traditional, static defenses at banks and credit unions struggle to keep up.”

Q: Can you tell us about Fideo Intelligence and what you do?

A: “At Fideo Intelligence, our identity graph links signals and uses that data to help our customers—primarily financial institutions and fintechs—detect, stop, monitor, and investigate fraud. We connect signals across channels so our customers can see relationships and patterns that individual data points alone cannot reveal.

Our focus is on bringing together a full picture of identity and behavior. Instead of looking at a device, a phone number, or an email address in isolation, we show how those signals relate to one another over time. That connected view is what lets our customers spot synthetic identities and AI-driven fraud workflows much earlier.”

“The connections between data points are often where the real story lives.”

Q: How does the concept of omnichannel fit into the fraud problem?

A: “Omnichannel is central to what we do, but probably not in the way people usually talk about it. Our graphs make sure customers can see the connections between signals across every channel—mobile, online, call center, branch, and more.

What is important to understand is that fraudsters do not think in terms of channels at all. ‘Channel’ is an internal construct that banks and credit unions use. Fraudsters think in terms of workflows and outcomes: ‘What is the next step in this sequence to get to an account takeover, a bust-out, or a money movement event?’ If you only think in channels, you are already a step behind.”

Q: Who’s actually behind large-scale fraud operations?

A: “These are large, sophisticated organizations—cartels, state actors, and groups engaged in terrorist financing. They operate with the same structure you would find in a legitimate company. They have OKRs, business goals, and large teams of people working in places like Cambodia, Laos, or Myanmar, conducting crimes and automating their workflows at scale.

When we talk about fraud today, we have to reckon with the fact that we are up against adversaries who are organized, well-funded, and incentivized to keep optimizing. They are not experimenting casually; they are running an ongoing operation that they expect to grow quarter after quarter.”

Q: How do these organizations approach targeting victims—is it methodical?

A: “It is extremely methodical. Fraud organizations think in terms of automated workflows. If a phishing attempt does not work, they do not give up. They pivot. Maybe they try a different message, a different channel, or a different method to compromise a device.

Once they succeed in compromising a device or an identity, they follow that workflow through to the end goal: an account takeover, moving money, building up credit to eventually bust out. Sometimes the goal is not even directly financial—it can be about creating disruption within financial institutions or overwhelming compliance functions. But in every case, the process is structured, repeatable, and designed to be improved over time.”

“The critical shift is to stop thinking of fraud as an event and start thinking of it as a sequence…. Fraud is a chain of steps, and you have to map all of those out together.”

Q. How should fraud-detection teams respond to this kind of adversary?

A. “The critical shift is to stop thinking of fraud as an event and start thinking of it as a sequence. You cannot just look at one signal—a phone, a device, an email—in isolation and expect to understand what is happening.

Fraud is a chain of steps, and you have to map all of those out together. That means connecting identity, device, and transaction signals into a single view, so you can see how changes and behaviors relate over time. When you do that, you start to see the workflows that professional fraud organizations are running, rather than just the individual alerts they generate.”

Q. What is graph intelligence and why does it matter?

A. “Graph intelligence is a way of analyzing relationships between data points—entities, behaviors, signals—by treating them as a network instead of a flat list. In fraud detection, that means linking identities, devices, accounts, and transactions in a graph rather than looking at each one in isolation.

When you graph these connections, patterns emerge: clusters of activity, shared infrastructure, and suspicious sequences. That is where you find anomalies. The connections between data points are often where the real story lives, especially when you are dealing with synthetic identities, mule networks, and coordinated AI-driven campaigns.”

Q. How should financial institutions balance fighting fraud and minimizing friction in the customer experience?

A. “There is a lot less friction now than there used to be, which is generally a good thing for customers. In some cases, financial institutions are actually allowing a little bit more fraud in and then catching it later. I am not saying that is the right thing to do, but that is what is happening in parts of the market.

The problem that creates is if you say, ‘We are going to catch it in the AML process,’ then the compliance piece gets a lot more cumbersome. If that is your strategy, you have to make sure those functions are very robust and very efficient. Otherwise, you are going to end up adding more people and more costs on the back end without fully solving the fraud problem.”

Fraud prevention shouldn’t come at the cost of customer experience. Connect with us here and learn how you can plug in Fideo and quickly detect what others miss.