ECPR

Install the app

Install this application on your home screen for quick and easy access when you’re on the go.

Just tap Share then “Add to Home Screen”

From Alerts to Intelligence: Analyst-Guided Mitigation of Organised E-Commerce Fraud Waves

Organised Crime
Internet
Decision Making
Technology
Artificial Intelligence
Egemen Bezci
University for Continuing Education Krems
Egemen Bezci
University for Continuing Education Krems

Friday 08:30 - 10:00 CET (11/12/2026) Building: Social Science Department, Room: II.2


Abstract

Financial fraud increasingly manifests not as isolated transactions, but as organised and scalable abuse across digital commerce infrastructures. Offenders exploit gaps between identity verification, payment systems, consumer behaviour, and merchant operations. Stolen credentials and coordinated chargeback activity can therefore be turned into high-frequency forms of financial crime. Although fraud prevention increasingly relies on automated risk scoring and machine-learning models, adaptive fraud tactics expose the limits of purely automated detection workflows. Drawing on an anonymised embedded case study within a major European financial technology firm, this paper examines how organised e-commerce fraud becomes visible inside high-volume risk operations. It frames fraud prevention as an intelligence-led socio-technical process. In this process, machine-generated alerts become operationally meaningful through analyst judgement, contextual knowledge, and documented intervention practices. Operationally, fraud analysts are not passive reviewers of algorithmic flags. Analysts connect fragmented transaction-level signals to broader patterns of abuse. These include repeated identity misuse, coordinated chargeback behaviour, shared behavioural indicators, and merchant-specific fraud waves. By interpreting weak signals across cases, analysts can identify emerging synthetic identity clusters before these patterns are sufficiently pronounced to trigger automated model thresholds. These judgements are then fed back into the decision system. The paper contributes to debates on organised financial crime, algorithmic governance, and responsible AI. It shows how analyst-guided systems can support earlier intervention and more defensible decision making in high volume fraud environments. It also argues that explainability, proportionality, and auditability are not only abstract governance principles, but practical features of well-designed fraud prevention workflows.