A Graph Neural Network Platform for Blockchain AntiMoney Laundering: A Literature Review and Proposed System Architecture

This paper reviewed eighteen studies on graph neural networks for blockchain anti-money laundering, spanning foundational Elliptic-benchmark models, attention-based and heterogeneous architectures, temporal and ensemble methods, explainability and privacy-preserving techniques, and early real-time system designs. The analysis shows that GNNs have become the dominant and consistently effective way to capture coordinated on-chain laundering behaviour. It also shows that the field has concentrated heavily on offline detection accuracy over a single blockchain and a single historical period, with comparatively little attention to incremental graph maintenance, inference latency, cross-chain generalization and architecture-native, investigator-facing explanation. To address these gaps, the proposed ChainSentry-GNN architecture combines chain-agnostic ledger ingestion, incremental graph updating, directed heterogeneous graph-attention scoring (Eqs. 5–7), reinforcement-learning-driven adaptive thresholding (Eqs. 9–10), typology-aligned explanation and a closed feedback loop with human investigators in a single platform design. By treating update latency, throughput and explanation fidelity as first-class evaluation criteria alongside detection accuracy, the proposed direction aims to help move blockchain AML research from static, benchmark-optimized models toward deployable, auditable monitoring platforms.

[1] M. Weber, G. Domeniconi, J. Chen, D. K. I. Weidele, C. Bellei, T. Robinson, and C. E. Leiserson, “Anti-Money Laundering in Bitcoin: Experimenting with Graph Convolutional Networks for Financial Forensics,” in Proc. KDD Workshop on Anomaly Detection in Finance, 2019. [2] C. Bellei, M. Xu, R. Phillips, T. Robinson, M. Weber, T. Kaler, C. E. Leiserson, Arvind, and J. Chen, “The Shape of Money Laundering: Subgraph Representation Learning on the Blockchain with the Elliptic2 Dataset,” in Proc. KDD Workshop on Machine Learning in Finance, 2024, arXiv:2404.19109. [3]  I. Alarab, S. Prakoonwit, and M. I. Nacer, “Competence of Graph Convolutional Networks for Anti-Money Laundering in Bitcoin Blockchain,” in Proc. 5th Int. Conf. Machine Learning Technologies (ICMLT), Beijing, China, 2020, pp. 23–27. [4]  I. Alarab and S. Prakoonwit, “Graph-Based LSTM for Anti-Money Laundering: Experimenting Temporal Graph Convolutional Network with Bitcoin Data,” Neural Process. Lett., vol. 54, pp. 1–19, 2022. [5] W. W. Lo, M. Sarhan, S. Layeghy, and M. Portmann, “Inspection-L: A Self-Supervised GNN-Based Money Laundering Detection System for Bitcoin,” arXiv:2203.10465, 2022. [6]  D. Cheng, Y. Ye, S. Xiang, Z. Ma, Y. Zhang, and C. Jiang, “Anti-Money Laundering by Group-Aware Deep Graph Learning,” IEEE Trans. Knowl. Data Eng., vol. 35, no. 12, pp. 12444–12457, 2023. [7] X. Li, Y. Li, X. Mo, H. Xiao, Y. Shen, and L. Chen, “DIGA: Guided Diffusion Model for Graph Recovery in Anti-Money Laundering,” in Proc. 29th ACM SIGKDD Conf. Knowledge Discovery and Data Mining, 2023, pp. 4404–4413. [8]  O. M. Olaniyi, I. S. Aroh, O. Henry, O. C. Metibemu, and O. I. Akinola, “Graph Neural Networks for Multi-Layered Financial Crime Network Detection: An Explainable AI Framework for Anti-Money Laundering,” J. Eng. Res. Rep., vol. 28, no. 2, pp. 18–36, 2026. [9] C.-H. Poon, J. Kwok, C. Chow, and J.-H. Choi, “LineMVGNN: Anti-Money Laundering with Line-Graph-Assisted Multi-View Graph Neural Networks,” AI, vol. 6, no. 4, art. 69, 2025. [10]  F. Johannessen and M. Jullum, “Finding Money Launderers Using Heterogeneous Graph Neural Networks,” arXiv:2307.13499, 2023.