News
Bitcoin forensic analysis reveals money laundering clusters and proceeds of crime
May 1, 2024Press roomFinancial Crime / Forensic Analysis
A forensic analysis of a graphical dataset containing transactions on the Bitcoin blockchain revealed clusters associated with illicit activity and money laundering, including the detection of criminal proceeds sent to a crypto exchange and wallets up to ‘then unknown people belonging to a Russian darknet market.
THE results come from Elliptic in collaboration with researchers from the MIT-IBM Watson AI Lab.
The 26 GB dataset, named Elliptical2is a “large graph dataset containing 122,000 labeled subgraphs of Bitcoin clusters in a background graph consisting of 49 million node clusters and 196 million edge transactions,” the co-authors said in an article shared with The Hacker News.
Elliptic2 relies on the Elliptical dataset (aka Elliptic1), a transaction chart made public in July 2019 with the aim of fight against financial crime using graphical convolutional neural networks (GCN).
The idea, in a nutshell, is to uncover illegal activities and money laundering schemes by leveraging blockchain pseudonymity and combining it with knowledge about the presence of legal (e.g. exchange, wallet provider, miner, etc.) and illicit (e.g. darknet market, malware, terrorist organizations, Ponzi scheme, etc.) services. on the network.
“Using machine learning at the subgraph level – that is, the groups of transactions that constitute money laundering cases – can be effective in predicting whether crypto transactions constitute money laundering. proceeds of crime,” said Tom Robinson, chief scientist and co-founder of Elliptic. Hacker news.
“It’s different from conventional crypto anti-money laundering (LBC), which rely on tracing funds from known illicit wallets or matching with known money laundering practices.
The study, which experimented with three different subgraph classification methods on Elliptic2, such as GNN segment, Sub2VecAnd GLASSidentified subgraphs that represented crypto exchange accounts potentially engaged in illegitimate activities.
On top of that, it helped trace the source of funds associated with suspicious subgraphs to various entities, including a cryptocurrency mixer, a Panama-based Ponzi scheme, and an invitation-only Russian dark web forum .
Robinson said that simply considering the “shape” – local structures within a complex network – of money laundering subgraphs has proven to be an already effective way of reporting criminal activity.
A closer look at the predicted subgraphs using the trained GLASS model also identified known patterns of cryptocurrency laundering, such as the presence of peeling chains and nested services.
“A peel chain is where a small amount of cryptocurrency is “peeled” to a destination address, while the rest is sent to another address under the user’s control,” Robinson explained. “This occurs repeatedly to form a peeling chain. This pattern may have legitimate financial privacy purposes, but it may also indicate money laundering, particularly when the “peeled” cryptocurrency is repeatedly sent to an exchange service.
“This is a known cryptocurrency laundering technique that bears an analogy to “smurfing” within traditional finance. So the fact that our machine learning mode independently identified it is encouraging.”
As for next steps, research should focus on increasing the accuracy and precision of these techniques, as well as extending the work to other blockchains, Robinson added.
Did you find this article interesting ? follow us on Twitter And LinkedIn to read more exclusive content we publish.
Fuente