Blockchain
Blockchain Sleuth Elliptic Explores AI and Anti-Money Laundering Using 200 Million Bitcoin Transactions
-
The Elliptic2 dataset is orders of magnitude larger than the one used when the team began using machine learning to detect bitcoin money laundering in 2019.
-
The research used 122,000 groups of connected nodes and transaction chains called “subgraphs” with known links to illicit activity.
Blockchain analytics firm Elliptic said it has detected potential money laundering patterns on the Bitcoin blockchain after training an artificial intelligence (AI) model using a record 200 million transactions.
The work is an extension of a realized program in 2019 which used a dataset of just 200,000 transactions. The much larger “Elliptic2” dataset used 122,000 labeled “subgraphs,” groups of connected nodes, and transaction chains known to have links to illicit activity.
AI becomes more in-depth the larger the dataset is available to train machine learning algorithms, and cryptocurrencies like bitcoin offer an abundant supply of transparent transaction data on the blockchain. Elliptic used the transactions to learn the set of “shapes” that money laundering presents in cryptocurrency and to accurately classify new criminal activities, Elliptic said in an article written in collaboration with researchers at the Watson AI Lab at MIT-IBM.
“The money laundering techniques identified by the model were identified because they are prevalent in bitcoin,” Elliptic co-founder Tom Robinson said in an email. “Crypto money laundering practices will evolve over time as they cease to be effective, but one advantage of an AI and deep learning approach is that new money laundering patterns are automatically identified as they emerge. “
Many of the suspicious subgraphs were found to contain so-called “peeling chains,” where a user sends or “peels” cryptocurrency to one destination address, while the rest is sent to another address under the user’s control. This happens repeatedly until a peeling chain is formed.
“In traditional finance this is known as ‘smurfing,’ where large amounts of money are structured into multiple small transactions, to keep them within regulatory reporting limits and avoid detection,” Elliptic said in the paper.
Another commonly used technique was the use of so-called “nested services,” businesses that move funds across accounts at larger cryptocurrency exchanges, sometimes without the exchange’s knowledge or approval. A nested service could receive a deposit from one of its customers into a cryptocurrency address and then forward the funds to its deposit address at an exchange.
“Nested services are known to often have less rigorous customer due diligence checks than the cryptocurrency exchanges they use, or sometimes have no such AML controls at all, resulting in their misuse for cryptocurrency laundering, potentially causing them to be found in subgraphs deemed by the model to be suspicious,” Elliptic said.