A collaboration between blockchain forensic company Elliptic, IBM Watson, and MIT researchers has led to significant progress in utilizing artificial intelligence (AI) for the detection of money laundering activities on the Bitcoin blockchain.
As per a recent report published by Elliptic on Wednesday, the researchers employed a deep learning model to effectively identify instances of cryptocurrency-related crimes. Titled “Enhancing Blockchain Analytics Through AI,” the study’s objective is to assist customers in more accurately evaluating risks associated with crypto assets.
The deep learning AI model has the capability to identify patterns of money laundering and pinpoint cryptocurrency wallets associated with criminal activities, according to the findings. Elliptic emphasized that unlike traditional financial systems, where transaction data is often compartmentalized, blockchain technology offers transparency, enabling the application of these techniques more effectively.
“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.”
In addition, Elliptic elaborated on how their machine learning model is specifically trained to detect “subgraphs,” which are chains of transactions known to signify Bitcoin money laundering activities. By focusing on these subgraphs, the model can analyze the broader “multi-hop” laundering process rather than solely examining the on-chain behavior of individual illicit actors. This approach enables a more comprehensive understanding of the laundering process within the Bitcoin blockchain ecosystem.
In simpler terms, the researchers utilized patterns observed in Bitcoin transactions originating from malicious actors and leading to cryptocurrency exchanges. These identified patterns were then used to train an AI model, which is capable of identifying similar money movements indicative of suspected money laundering behavior.
Furthermore, Elliptic emphasized that they employed these advanced AI tools to analyze a vast dataset comprising over 200 million transactions. This extensive dataset allowed for thorough testing and validation of the AI model’s effectiveness in identifying and flagging potential instances of money laundering on the Bitcoin blockchain.
Money laundering cast a shadow over the cryptocurrency landscape in 2023.
In its 14th year of existence, the cryptocurrency realm in 2023 witnessed a concerning trend, with an estimated $22.2 billion worth of cryptocurrency being laundered, as outlined in Chainalysis’s report.
Despite this staggering figure, there was a notable decline from the previous year, with 2022 recording $31.5 billion in laundered funds. Encouragingly, the reduction in money laundering activity outpaced the drop in total transaction volume, with a decline of 29.5% compared to a 14.9% decrease in overall transaction volume.
It’s worth highlighting the evolving tactics employed by crypto launderers, with many sophisticated criminals now leveraging bridges and mixers to obfuscate the origins of illicitly obtained funds.
In response to these challenges, governments worldwide have ramped up efforts to enhance vigilance and enforce anti-money laundering (AML) regulations. This heightened scrutiny has resulted in significant legal actions against prominent figures within the cryptocurrency industry.
For instance, just this week, former Binance CEO Changpeng Zhao was sentenced to four months in federal prison for money laundering violations. Zhao’s guilty plea in 2023 marked a pivotal moment for Binance, the world’s largest cryptocurrency exchange.
Similarly, in another high-profile case last month, US authorities arrested the founders of Bitcoin mixer Samourai on charges of money laundering. The founders stand accused of facilitating over $2 billion in illicit transactions and laundering in excess of $100 million in criminal proceeds.
