The research used 122,000 'subgraphs' with known links to illicit activity, detecting money laundering patterns after training AI with a record 200 million transactions.
Elliptic's AI model learns the 'shapes' of money laundering in cryptocurrency from transparent transactions on the blockchain, constantly evolving to identify new criminal activity automatically.
Elliptic co-founder Tom Robinson highlights the evolution of crypto laundering techniques and the advantage of AI in automatically identifying new patterns as they emerge.
Suspicious subgraphs often involved 'peeling chains,' where cryptocurrency is sent to one address, and the remainder to another under the user's control.
[
add
]
[
|
|
...
]