2025 is the year of adaptive fraud prevention, say experts

In horror movie terms, fraud is a shapeshifting monster. In the era of easily accessible generative AI tools, fraud morphs at the pace of criminal ingenuity amplified at massive scale. Some say static fraud prevention methods are no longer adequate – that fraud prevention, too, must operate in a state of constant adaptation, spotting new threats in real time.

National fraud strategy must prioritize prevention, accountability: BioCatch

UK Finance has released its annual fraud report, in partnership with BioCatch. And the news is consistent, if bad: “fraud continues to be the most common crime in the UK,” it says. While the fraud landscape changes over time, with some techniques proving more popular or effective, the cumulative effect is still massive: per the report, “the net effect was over £1.1 billion of losses in 2024.”

Authorized push payment (APP) fraud, in which victims are tricked into making payments, is down. However, remote purchase fraud has increased to take up the slack.

The agility of fraudsters means that fraud prevention must be seen as a process, rather than an end result. “Rather than trying to ‘solve’ fraud, our objective should be to protect as many people as possible by reducing and managing the evolving threat,” says an introduction from Ben Donaldson, managing director of economic crime for UK Finance. He calls for “a system-wide, strategic approach to countering fraud, rather than tactical interventions which target individual components of the broader threat.”

Labeling the fraud issue a matter of national security, Donaldson says “our collective priority should be preventing these crimes from happening in the first place. We need layers of defence spanning the technology, telecommunications, financial services, and public sectors.”

Director of Global Advisory for EMEA at BioCatch Jonathan Frost says that, while industry is holding its own against the rising tide of fraud, “the gains made by fraudsters demonstrate that the fraud threat is persistent; tackling it necessitates that we collectively innovate and adapt at a pace they cannot match.”

When one hole is plugged, another emerges: investment scams are on the rise, as the UK government considers measures to encourage consumers to switch their cash savings into stocks and shares, and remote purchase fraud could get supercharged by a new development in the upcoming iPhone that allows for remote control.

Frost also believes that fighting the surge requires the development of a strong comprehensive national fraud strategy.

“Outdated policing, patchy regulation across sectors like crypto and telecoms, and reimbursement rules have created an over-reliance on banks,” he says. “If the government is serious about stopping this tsunami of digital fraud, its national fraud strategy must prioritize prevention, demand accountability from all sectors, embrace smarter technologies and drive real collaboration across law enforcement, banks, tech firms and telecoms.”

“Until then, banks will remain unfairly burdened, while fraudsters continue to profit from a system that inevitably fails too many.”

Power of AI only unlocked through access to meaningful, quality data: Feedzai

A release from Feedzai announces the launch of Feedzai IQ, a new family of products that “enables network fraud intelligence without compromising customer privacy.” The tools leverage insights from hundreds of financial institutions across four continents and over $8 trillion in annual payment volume.

“We’ve always believed that the true power of AI is only unlocked through access to meaningful, high-quality data,” says Pedro Barata, chief product officer at Feedzai. “While AI is surrounded by hype today, Feedzai has led the way in applying real AI to real problems – and now, with Feedzai IQ, we’re combining our AI expertise with secure, network-wide intelligence.”

Feedzai IQ uses federated learning to transform anonymized, distributed data into real-time fraud intelligence. Its features include real-time fraud risk scoring and pre-calculated risk indicators to boost detection accuracy and improve payment acceptance.

Goal is to create ‘cyber-DNA’ based derived from behavioral biometrics: Facephi

Spain’s Facephi also has a new real-time fraud prevention product. A release says its Behavioral Biometrics tech analyzes over 3,000 contextual signals and is capable of generating a unique behavioral profile for each user to detect suspicious activity – “without relying on traditional biometrics.” Instead, it identifies individual characteristics in how people interact with their electronic devices: changes in typing rhythm, screen focus shifts, abnormal response times or “details about when, from where, and what the user is accessing.”

The goal, says Facephi General Manager Jorge Sanz, “is to create a sort of ‘cyber-DNA’ based on various user parameters derived from behavioral biometrics – something as unique as their fingerprints.”

By evaluating every digital interaction within its context, Facephi ensures that only legitimate users gain access to systems and services in highly regulated sectors such as banking, fintech, insurance and healthcare.

Fraud teams must move beyond static defenses: SEON

SEON has launched its expanded Anti-Money Laundering (AML) Compliance suite, calling it in a release “a significant step in the company’s evolution from fraud prevention to a unified risk solution.” It introduces real-time payment screening, transaction monitoring and integrated fraud and AML case management with regulatory reporting functionality.

“Risk teams don’t need more tools – they need one that gives them a full picture,” says Tamás Kádár, CEO of SEON. “This launch gives fraud and compliance teams a multi-dimensional command center to manage risk in one place. And because we own our data pipeline and can control data accuracy, this means our models deliver more accurate risk decisions.”

The tool enables teams to investigate efficiently and close out cases faster with unified fraud intelligence, AML signals and transaction data in one dashboard; reduce time spent on regulatory reporting with autofill capabilities and AI-powered SAR narratives; and submit SAR, CTR and Form 8300 reports to FinCEN in one click, with status tracking and audit history.

According to SEON’s 2025 Digital Fraud Outlook, real-time transaction monitoring is the top investment priority for compliance and fraud teams this year, with 62 percent of organizations making the shift. “Fraud teams must move beyond static defenses to dynamic, contextaware models that continuously adapt in real time,” Kádár says. “This shift will enable organizations to anticipate threats, respond instantly and refine fraud strategies as risks evolve.”

Summy is there to support every step of the way: Sumsub

Sumsub has announced a major update to its Case Management solution, which a release calls “a ready-to-use, cross-functional hub for financial crime investigations designed to streamline fragmented risk operations.”

The name of the game is the same: automation and real-time risk assessment. But that’s not all: the upgrade also introduces Summy, Sumsub’s AI assistant, to provide case summaries and action recommendations for risk, compliance and fraud teams – sort of like a highly evolved Clippy for Case Management.

“Fragmentation is one of the biggest threats in modern risk management,” says Andrew Novoselsky, chief product officer at Sumsub. “With the rise of deepfakes and complex fraud schemes, teams can’t afford to work in silos. That’s why we built a platform that’s not just unified, but smart – where AI like Summy supports every step of the investigation.”

Article Topics

AI fraud  |  BioCatch  |  biometrics  |  digital identity  |  FacePhi  |  Feedzai  |  fraud prevention  |  generative AI  |  identity verification  |  SEON  |  Sumsub