AI-Powered Money Laundering Detection on the Bitcoin Blockchain
Researchers Use Machine Learning to Identify Illicit Activity
Researchers from Elliptic, IBM Watson, and MIT have made a significant breakthrough in the fight against money laundering on the Bitcoin blockchain. By applying artificial intelligence (AI) techniques, they have developed a system that can detect and identify illicit activity, including money laundering, with a high degree of accuracy.
Background
In 2019, Elliptic, a blockchain analytics firm, published research with the MIT-IBM Watson AI Lab on training a machine learning model to identify Bitcoin transactions made by illicit actors, such as ransomware groups or darknet marketplaces. This initial research laid the foundation for further exploration and development of AI-powered money laundering detection techniques.
New Research and Methodology
In their latest research, the partners have applied new techniques to a much larger dataset, containing nearly 200 million transactions. Instead of identifying transactions made by illicit actors, the machine learning model was trained to identify “subgraphs,” chains of transactions that represent Bitcoin being laundered.
This approach allows the researchers to focus on the “multi-hop” laundering process more generally, rather than the on-chain behavior of specific illicit actors. By identifying these subgraphs, the system can detect and prevent money laundering, even if the illicit actors are using sophisticated methods to conceal their activities.
Testing and Results
The researchers worked with a crypto exchange to test their technique. They predicted 52 money laundering subgraphs, and 14 of these ended with deposits to the exchange. Of these 14 subgraphs, 14 were received by users who had already been flagged as being linked to money laundering.
The team reported that, on average, less than one in 10,000 accounts are flagged in this way, suggesting that the model performs very well. The researchers are now making their underlying data publicly available, allowing other experts to build upon their findings and develop even more effective AI-powered money laundering detection systems.
Conclusion
The use of AI-powered money laundering detection on the Bitcoin blockchain has significant implications for the financial industry. By leveraging the transparency of blockchain data, researchers can develop more effective methods for identifying and preventing illicit activity, ultimately making it more difficult for criminals to launder money and finance illegal activities.
FAQs
Q: What is the significance of this research?
A: This research demonstrates the potential of AI-powered money laundering detection on the Bitcoin blockchain, showcasing the ability to identify and prevent illicit activity with a high degree of accuracy.
Q: How does the system work?
A: The system uses machine learning to identify “subgraphs,” chains of transactions that represent Bitcoin being laundered. This approach allows the system to focus on the “multi-hop” laundering process more generally, rather than the on-chain behavior of specific illicit actors.
Q: How accurate is the system?
A: According to the researchers, the system performs very well, with less than one in 10,000 accounts flagged as being linked to money laundering. This suggests a high degree of accuracy in identifying illicit activity.
Q: What are the implications for the financial industry?
A: The use of AI-powered money laundering detection on the Bitcoin blockchain has significant implications for the financial industry, allowing for more effective methods for identifying and preventing illicit activity and ultimately making it more difficult for criminals to launder money and finance illegal activities.
Q: Will this technology be publicly available?
A: Yes, the researchers are making their underlying data publicly available, allowing other experts to build upon their findings and develop even more effective AI-powered money laundering detection systems.
