The ever-evolving threat of fraud and financial crime

Orchestration bridging the gaps in fraud prevention:

The first major evolution is the maturation of information orchestration into a mandatory, real time mechanism, particularly to combat AI-driven fraud. Banks will move past static, point in time checks. Instead, real time data feeds, such as continuous mule list ingestion from external third parties (cyber, credit, etc) and credit risk signals will be orchestrated at the precise point of payment for certain high risk transactions. Orchestration could also be used to initiate document checks to detect AI manipulation on account opening or even at point of inbound/outbound friction as proof of entitlement or movement of funds. This fusion allows institutions to instantly correlate a transaction with a high risk account number or an immediate spike in credit utilisation, providing the holistic visibility needed to intercept complex synthetic identity fraud for example.

‍

Hyper-granularisation of fraud profiles:

The industry is abandoning the ‘one size fits all’ approach to scams. 2026 will see the widespread adoption of multi-model environments where specialist AI is deployed to profile the type of attack, not just the customer. For instance, a bespoke ‘romance scam’ model targets specific behavioral red flags (e.g., frequent, rapid transfers to an international destination), while a separate ‘investment scam’ model focuses on velocity spikes to newly established accounts. This specialisation promises to dramatically lower false positive rates and increase detection accuracy against socially engineered attacks. It also facilitates more contextualized customer alerts that have a much higher chance of being heeded than repeated generic warnings.

‍

Proliferation of mobile wallet fraud across both third party and first party vectors:

As real time payments accelerate, mobile wallets become prime targets. While sophisticated account takeover attacks remain a threat, the key trend for 2026 will be the surge in first party fraud, where genuine customers engage in activity like friendly fraud. PSD3 will enforce strong customer authentication here and ideally require in-app checks to provision wallets so that device/location intelligence can be extracted and used in identifying anomalous transactions.

‍

Greater international and law enforcement involvement:

This will focus on dismantling the infrastructure of crime. We will see intensified political and operational pressure aimed at shutting down large scale ‘scam centres’. This concerted global effort, coupled with improved cross-border data sharing, will disrupt the money laundering networks and mule recruitment tactics, forcing criminal enterprises to become more fragmented and agile, rather than centralised and industrialised.

‍

Necessary friction for customers:

Institutions will adopt a much tougher stance, including less benefit of the doubt for potential mule accounts, often resulting in account restriction or closure based on strong risk indicators. Furthermore, customer experience will be traded for security through operational friction, such as blocking the ability to initiate large transfers if customers are detected to be on an active, external phone call, a key indicator of a potential social engineering scam. More stringent, step up authentication journeys will become standard for high risk, high value payments.

‍

The fraud fight in 2026 demands a unified, real time risk platform. A sustainable defense is achieved through the orchestration of maximum data intelligence and hyper-granular models, which automatically counter rapidly evolving threats and intelligently manage the implementation of necessary friction to protect consumers, turning external pressure into operational advantage.