Bitcoin Forensic Analysis Uncovers Money Laundering Clusters and Criminal Proceeds

Elliptic, with MIT-IBM Watson AI Lab researchers, released Elliptic2 — a 26 GB labeled Bitcoin transaction graph dataset containing 122K subgraphs within a 49M-node background graph — and used GNN-based subgraph classification (GNN-Seg, Sub2Vec, GLASS) to detect clusters associated with illicit activity and money laundering, tracing proceeds to a crypto exchange, a mixer, a Panama Ponzi scheme, and a Russian darknet forum, and identifying laundering patterns such as peeling chains and nested services.