European banks confront AI, crypto-driven fraud complexity

A survey of 200 European fraud leaders conducted by Finextra in association with NICE Actimize found growing fraud complexity driven by AI-initiated payment flows and scams linked to virtual assets. The survey formed the basis of a 2026 webinar that brought together industry practitioners to discuss current fraud patterns and control responses across Europe.

The webinar included Chris Ainsley, Head of Fraud Risk Management at Santander UK; Joe Bristow, Product Director and fraud subject-matter expert at NICE Actimize; and Sharon Kimathi, researcher at Finextra, who moderated the session. Organisers said the survey and webinar aimed to map threats and identify operational priorities for fraud, risk and compliance teams.

Respondents reported that traditional fraud types such as card fraud and account takeover remain active while newer threats are emerging. The survey identified AI-driven payment authorisations and scams exploiting virtual-asset platforms as rising concerns that extend attack surfaces beyond retail banking into faster and less regulated payment rails.

Banks reported investments in machine-learning models and other AI tools to expand detection and response across established and emerging fraud types. Survey participants cited AI-based monitoring as a method to scale oversight and detect patterns that rule-based systems miss.

Many institutions reported barriers that slow deployment of new tools. Common issues included integration with legacy systems, lack of standardised data, and shortages of staff with relevant skills. These constraints were described as factors that delay moving AI-based controls into production environments.

The survey also found uneven preparedness across European markets. Smaller banks and regional firms were identified as having gaps in expertise, funding and access to advanced tools, while larger institutions were described as more able to maintain dedicated fraud teams and in-house analytics. Participants said these disparities affect how quickly firms can adopt controls and share intelligence on new threats.

Panel discussion priorities included improved telemetry across channels, stronger linkage of payments and identity signals, and enhanced monitoring for activity tied to virtual-asset flows. Webinar contributors outlined practical work streams for integrating detection tools, updating governance and reallocating resources to support new technology rollouts.

Organisers reported that the session aimed to help fraud, risk and compliance leaders set short-term priorities and to specify the technical work required to deploy AI-based controls across different banking environments. The webinar materials and survey data were presented as a basis for identifying immediate control gaps and next steps for implementation.