Elliptic, a blockchain analysis firm, released a study conducted with the MIT-IBM Watson AI Lab on Wednesday. This study utilized a deep learning AI model to identify money laundering patterns within Bitcoin transactions and pinpoint wallets involved in criminal activities.
Bitcoin’s use of a decentralized public ledger, a fundamental aspect of the technology, facilitated the research, according to Elliptic. The collaboration between Elliptic and MIT-IBM leveraged AI to distinguish between legal and illegal transactions, tracking connections within the latter to expose potential money laundering activities.
Elliptic stated, “Blockchains provide fertile ground for machine learning techniques, thanks to the availability of both transaction data and information on the types of entities that are transacting, collected by us and others. This is in contrast to traditional finance where transaction data is typically siloed, making it challenging to apply these techniques.”
Elliptic clarified that instead of directly identifying transactions by illicit actors, the machine learning model is trained to recognize ‘subgraphs’—sequences of transactions that depict the laundering of Bitcoin. This method focuses more broadly on the multi-step laundering process rather than specific illegal actions on the blockchain.
See Related: Binance’s $3.9 Billion USDT Transfer Sparks Speculation Amidst CEO’s Guilty Plea
Money Laundering in 2024 Reports
In 2023, the volume of cryptocurrency involved in money laundering activities decreased significantly, with centralized exchanges continuing as the primary channels for such transactions. DeFi protocols and gambling services saw an uptick in use for laundering, yet they generally lack the anonymity preferred by launderers.
Sophisticated criminals, particularly North Korea’s Lazarus Group, adapted to law enforcement tactics by employing new mixers and utilizing cross-chain bridges to launder funds. These methods allow them to obfuscate the origins of illicit funds more effectively across multiple blockchain platforms, demonstrating a dynamic shift in the techniques used by high-level criminal enterprises to evade detection.
