Artificial intelligence (AI) technology has taken a step forward in the fight against money laundering on the Bitcoin blockchain, according to a recent report from Elliptic, a blockchain forensic firm. Collaborating with IBM Watson and MIT, researchers successfully used a deep learning model to detect crypto crimes. The report, titled “Enhancing Blockchain Analytics Through AI,” aims to provide customers with more accurate risk assessments of crypto assets.
The deep learning AI model is able to identify money laundering patterns and recognize crypto wallets used in illicit activities. Unlike traditional finance, where transaction data is often isolated, the transparency of the blockchain allows for the application of these techniques. The researchers trained a machine learning model to detect “subgraphs,” which are chains of transactions associated with Bitcoin money laundering. This approach allows them to focus on the overall process of money laundering, rather than the specific behavior of individual actors.
To train the AI model, the researchers analyzed patterns of Bitcoin transactions that originated from bad actors and led to crypto exchanges. They then used these patterns as examples to identify similar money movements indicative of suspected money laundering.
Elliptic tested the AI tools on a dataset containing over 200 million transactions. The report also highlighted the significant decrease in cryptocurrency laundering in 2023 compared to the previous year. According to Chainalysis’s report, $22.2 billion worth of cryptocurrency was laundered in 2023, a decrease from the $31.5 billion recorded in 2022. Notably, the drop in money laundering activity was steeper, at 29.5%, compared to the 14.9% drop in total transaction volume.
The tactics used for crypto laundering have evolved, with sophisticated criminals now relying on bridges and mixers. In response, governments have placed increased emphasis on the importance of adhering to anti-money laundering (AML) rules. Recent high-profile cases, such as the sentencing of Binance’s ex-CEO Changpeng Zhao to four months in federal prison for money laundering violations, highlight the growing crackdown on illicit activities in the crypto industry. In another case, the founders of Bitcoin mixer Samourai were arrested last month on money laundering charges, accused of executing over $2 billion in unlawful transactions and laundering over $100 million in criminal proceeds.
The progress made by Elliptic, IBM Watson, and MIT in using AI to detect money laundering on the Bitcoin blockchain marks a significant step forward in combating financial crimes in the crypto space.
