Chinese police academy developed AI framework detecting illicit Bitcoin transactions with 90% accuracy.
Regulation & Gov ·
Researchers at China's People's Public Security University, which operates under the Ministry of Public Security, have created an AI system capable of identifying illicit Bitcoin transactions at a 90% accuracy rate. The framework integrates a memory module and large language model to flag anonymous cross-border cryptocurrency trades linked to money laundering and other financial crimes, addressing vulnerabilities created by the pseudonymous and transnational characteristics of blockchain activity.
The development aligns with Beijing's intensifying enforcement posture on crypto-related criminal activity. In March, China's Supreme People's Procuratorate disclosed that prosecutors had brought charges against 3,259 individuals throughout 2025 in cases involving virtual currency money laundering and underground banking operations. Researchers framed the AI tool as supplying regulatory bodies with a new technological approach to combating illicit crypto transactions and economic offenses.
The study, published in May, does not disclose the scope of the training dataset, transaction volume tested, or whether the framework has been operationalized by law enforcement. The extent to which this detection capability may affect legitimate privacy-preserving cryptocurrency users or decentralized finance remains unclear.