European banks lag as AI and crypto frauds rise

In 2026 a survey of 200 European fraud leaders conducted for a webinar found banks are having difficulty keeping up with rising fraud complexity. Threats reported include traditional card fraud and account takeover as well as AI-initiated payment instructions and schemes targeting virtual-asset platforms.

Respondents said fraud is accelerating across multiple channels and that institutions face challenges against both established and new attack types. Automated payment instructions driven by artificial intelligence and fraud on virtual asset platforms were listed among the emerging concerns, while card fraud and account takeover remain common.

The webinar used the survey data to examine how banks are adjusting controls and where they are allocating investment. Fraud, risk and compliance teams across Europe are prioritizing AI-based detection and automation to expand coverage of existing fraud patterns and to identify new behaviors faster.

Many banks reported that integrating new technologies into existing systems and workflows is a significant obstacle. Survey participants cited legacy IT architectures, limited interoperability between vendor solutions, and constrained budgets for fraud teams as barriers to rapid modernization.

Those operational pressures delay the rollout and tuning of advanced controls even as fraud actors use tactics that span multiple channels. Respondents reported that AI-driven attacks require quicker model updates and more frequent reassessment of rule sets than many institutions can sustain with current resources.

The survey showed uneven levels of expertise, funding and tooling across European markets. Larger banks and firms in better-resourced markets have deployed machine-learning systems and cloud-native analytics to supplement rule-based controls. Other institutions rely mainly on manual reviews and older detection systems.

Investment priorities identified by participants include extending AI capabilities, improving data quality and integrating disparate fraud systems. Banks are exploring machine-learning models that analyze payment flows end-to-end and detect anomalies that span accounts, devices and external crypto platforms.

Several respondents highlighted the need for stronger telemetry and a single view of customer and transaction behavior so alerts can be correlated and incidents resolved more quickly. The webinar discussion also covered expanded real-time scoring, use of behavioral biometrics to detect account takeover, and strengthened monitoring of on- and off-ramp activity tied to virtual assets.

Panelists addressed regulatory alignment and cross-institution information sharing as responses to fraud that leverages multiple platforms and jurisdictions. They noted that disparities in tools and funding affect the ability of banks to tune controls and respond to coordinated threats.

Sharon Kimathi, researcher and the webinar moderator, reviewed the survey findings with the panel and guided discussions on where banks are directing investment and which operational barriers they report as they update fraud controls for 2026 and beyond.