Blockchain analytics company Elliptic and MIT-IBM Watson AI Lab have come together to integrate artificial intelligence technology in discovering Bitcoin money laundering. By leveraging deep learning, the model identified illegal profits, fresh laundering transaction trends, and previously undisclosed illicit wallets. These revelations are currently being utilized to enhance Elliptic's offerings.

Unveiling Concealed Money Laundering Patterns

Elliptic has publicly released a substantial dataset of more than 200 million transactions to empower the community in creating novel AI techniques for spotting unlawful cryptocurrency operations. Rather than concentrating on transactions carried out by criminals, a machine learning system is educated to recognize "subgraphs" - sequences of transactions that suggest Bitcoin laundering. This strategy enables Elliptic to center on the wider "multi-hop" laundering process instead of the specific activities of single wrongdoers.

Evaluating the Model's Effectiveness

In an undisclosed cryptocurrency exchange, Elliptic assessed the model's capability to identify money laundering activities. Among 52 predicted "money laundering" subgraphs concluding with deposits to this exchange, 14 were authenticated by the exchange as linked to highlighted users. The fact that less than one in 10,000 accounts are flagged indicates the model's impressive performance. The company believes there is substantial potential in this approach to unveil and deter illicit activities.