Ivaylo Bozoukov - The Rise of Financial Crime in the Digital Age

Financial crime has always adapted, but what’s different now is the speed.

Fraud attacks have not just grown in volume over the past two years. They have mutated. The tools available to criminals today, many of them powered by artificial intelligence, have fundamentally changed what a fraud attempt looks like, how long it takes to execute, and how difficult it is to detect . For the institutions tasked with stopping it, the rulebook is being rewritten in real time.

When Technology Alone Is Not Enough

For a long time, the assumption was that better technology would win the fight. Build smarter detection systems, tighten the controls, and fraudsters would run out of road. That logic no longer holds.

Between 2024 and 2025, cases involving social manipulation rose by 33% . Fraudsters are no longer firing off mass phishing attempts and hoping something sticks. They invest days, sometimes weeks, building trust with their targets. They mirror language, exploit emotional vulnerabilities, and guide victims towards harmful decisions so gradually that the victim is often unaware until the money has gone.

Ivo Bozukov puts the challenge plainly. " What's changed is that fraud now exploits human behaviour as reliably as it exploits technical weaknesses. Those two things have become equally dangerous, and most institutions are still only set up to fight one of them. "

The implications are uncomfortable. A transaction that passes every automated check, initiated by the legitimate account holder from their usual device and location, can still be fraudulent. The persuasion happened off - platform, and the technical controls never had a chance to catch it.

AI on Both Sides of the Problem

Artificial intelligence has accelerated this dynamic considerably. Synthetic identities, deepfake video for liveness checks, voice cloning, and fabricated documents are no longer niche capabilities available only to well -resourced criminal networks. They are increasingly accessible, increasingly convincing, and increasingly difficult to detect at onboarding.

Research from Experian found that AI enabled synthetic identities now account for 42% of third-party identity fraud cases. Once a synthetic identity clears one platform's verification, it can be reused across multiple payment firms with small variations. The attack scales, and the damage multiplies.

At the same time, financial institutions are deploying AI to fight back, using machine learning to detect anomalous transaction patterns, behavioural biometrics to flag account takeovers, and real-time decisioning systems that can score a transaction in milliseconds. The challenge is that deploying these tools without proper governance creates its own risks. Systems are only as reliable as the data fed into them, and where data integrity is poor, AI amplifies weakness as readily as it enhances strength.

Ivaylo Bozoukov is direct about where the limits lie. " Technology is essential, but it cannot operate in isolation. The institutions genuinely ahead right now are the ones that have built governance frameworks around their AI tools, not just switched them on and hoped for the best."

The Structural Shifts That Will Define the Response

Beyond the individual attack vectors, there are larger structural changes underway that will determine how effectively the industry responds.

One of the most consequential is the convergence of fraud, anti-money laundering, and cybersecurity functions . For years these operated as separate disciplines with different reporting lines, different tooling, and limited communication between teams. That separation no longer reflects operational reality. A phishing attack leads to credential compromise. Credential compromise enables account takeover. Funds are moved, layered, and dispersed across multiple payment rails, sometimes before AML systems register anything unusual. Individually , each team sees a fragment of the risk. Together, they see the attack chain. The firms investing in integrated visibility are the ones catching these incidents earlier.

The regulatory environment is shifting in the same direction. The era of batch reporting and retrospective review is giving way to real-time monitoring and earlier intervention. Regulators are no longer asking simply whether a firm complied with the rules. They are asking whether the firm could have spotted the pattern sooner, and whether it acted when it did.

Cross-sector collaboration is the other piece. A significant proportion of authorised push payment fraud originates outside the financial system entirely, through online platforms and telecommunications networks. UK Finance data suggests around 70% of APP fraud cases begin online, with a further 16% enabled via telecoms. Payment firms absorb a disproportionate share of the liability for scams they had no realistic opportunity to prevent upstream. Addressing that imbalance requires accountability to travel further up the chain.

"Financial crime at this scale cannot be solved by any single institution acting alone," says Ivo Bozukov. "The criminal networks are joined up. The response needs to be too."

The Road Ahead

The trajectory is clear enough. Fraud will keep growing in sophistication. The attack methods that feel novel today will be standard within eighteen months. The organisations that fare best will not necessarily be those with the largest compliance budgets or the most advanced tooling. They will be those that have built resilient control frameworks, invested in human expertise alongside technology, and positioned themselves to intervene early rather than investigate late.