An Agent Based Anti-Money Laundering System Architecture for Financial Supervision

Anti-money laundering is a hot issue nowadays. How to establish an automatically technology to extract SARs (suspicious activity report) is the core problem for financial institutions. This paper aims to establish system architecture based on agent technology financial supervision in financial institution taking into consideration of not only transaction information database in financial institutions, but also customer profiles information and external information. The capability of self-learning and self-adaptive makes it difficult for those customers to avoid supervision through simple learning from regulations.