Abstract

The proliferation of cryptocurrency markets in Nigeria presents a complex challenge for national security. It offers new avenues for terrorist organizations to obscure financial transactions. Traditional regulatory and f inancial surveillance methods are ill-equipped to analyze the pseudo-anonymous, high-volume, and non-lin ear transaction graphs inherent in blockchain economies. This pioneering study proposes a novel AI-driven framework to assess terrorism financing (TF) risks. The study utilized Graph Neural Networks (GNNs) to mod el the Nigerian cryptocurrency transaction landscape, mapping flow patterns and identifying latent network structures. Superimposed on this graph, an ensemble of unsupervised anomaly detection models, including Isolation Forests and Autoencoders, is deployed to flag high-risk transaction clusters and behavioral outliers. Our research pioneers a method to move beyond simplistic transaction monitoring to a holistic network-level risk assessment. The findings demonstrate AI's capacity to deconstruct emerging TF typologies in real-time, offering a paradigm shift from reactive compliance to proactive intelligence-led disruption. We conclude by critically evaluating this AI framework against the nascent regulatory responses in Nigeria, proposing a syn ergistic model where adaptive AI tools can inform and future-proof financial policy.

DOI: doi.org/10.63721/25JPAIR0116

To Read or Download the Article  PDF