TechFlow News, May 18: According to official announcements, Bybit recently completed an innovative anti-money laundering (AML) research collaboration with a student team from The University of Hong Kong (HKU). Using Bybit’s February 2025 security incident as a case study, this collaboration provided HKU Business School master’s students with hands-on experience in blockchain investigations and AML analysis.
This capstone project was led by Assistant Professor Doyeon Kim of HKU’s Department of Accounting and Law, with full guidance from Bybit’s risk control and security expert David Zong. The student team simulated the tracing of fund flows linked to the North Korean hacker group Lazarus Group and developed machine learning models to identify coin mixer transaction patterns and suspicious activities.
The student team analyzed approximately 49,800 Bitcoin blocks and over 146 million transactions, applying clustering models and graph neural network techniques for in-depth analysis. Their research identified 10,289 transactions resembling those of Wasabi, constructing a blockchain subgraph comprising over 1.6 million address nodes and nearly 6 million transaction edges. One of the machine learning clusters achieved a recall rate of 70.5% for confirmed North Korea–linked addresses.
Students participating in the project stated that this experience significantly enhanced their understanding of blockchain investigations, AML systems, and cryptocurrency security operations. It deepened their appreciation of how AML technologies raise the operational barriers for criminals while also exposing them to the complexities involved in solving real-world industry challenges. Several students noted that this collaboration offered invaluable perspectives for their future careers in blockchain security, compliance, and finance.
