国际清算银行-Hertha项目：识别实时零售支付系统中的金融犯罪模式（英）

Identifying financial crime patterns in real-time retail payment systemsProject HerthaContentsBIS Innovation Hub Project Herthabis.org1Executive summary1. Motivation and hypotheses2. Results3. Key insights4. Areas for further researchExecutive summary 1. Motivation and hypotheses 2. Results 3. Key insights 4. Areas for further research Executive summarySynthetic data setThe experiments were conducted using a complex simulated synthetic transaction data set, developed as part of the project. It includes data on 1.8 million bank accounts and 308 million transactions. The data set was built using an AI model trained to simulate realistic transaction patterns. While no real customer data was used in the exercise, the data set was designed to be representative of an ecosystem of retail payments in a single jurisdiction. Project Hertha is a joint project between the BIS Innovation Hub’s London Centre and the Bank of England. The project explored how transaction analytics could help identify financial crime patterns in real-time retail payment systems, while using the minimum set of data points.MotivationCombatting financial crime is essential to maintaining trust in the financial system. It has been estimated1 that $3 trillion of money laundering and terrorist financing flow through the global financial system every year. Addressing this is increasingly urgent as new technologies are also enabling new financial crime threats.To evade detection, criminals operate in complex networks which include many accounts across multiple financial institutions. Earlier initiatives, including the BIS Innovation Hub’s Project Aurora, demonstrated the potential of network analytics to identify this activity in network-wide data. Electronic payment systems process transactions across many participants, which gives them a network-wide view. Project Hertha tested the application of modern artificial intelligence (AI) techniques to help spot complex and coordinated criminal activity in payment system data. It measured the added value of such transaction analytics relative to a modelled benchmark of banks and payment service providers (PSPs) monitoring accounts in isolation. 1 Nasdaq Verafin, Global financial crime report, 2024.BIS Innovation Hub Project Herthabis.org21. Motivation and hypotheses2. Results3. Key insights4. Areas for further researchExecutive summaryThe results have been achieved while using a minimal number of data points, demonstrating that advanced models can draw on network patterns rather than personal data.Executive summaryThe results demonstrate promise, but also show there are limits to the application and effectiveness of system analytics. It is just one piece of the puzzle. The introduction of a similar solution would also raise complex practical, legal and regulatory issues. Analysing these was beyond the scope of Project Hertha. The concept explored in the project does not assume any changes in the responsibilities of individual ins