Skip to the main content.

1 min read

Data Science and Artificial Intelligence Enhance Correspondent Banking Transaction Monitoring

Two key challenges within large banks for transaction monitoring investigations include establishing an economic purpose and verifying complementary lines of business for correspondent banking (CB) customers. Banks often find themselves filing conservative suspicious activity/transaction reports (SARs) because the necessary data is unavailable. Fortunately, data science and artificial intelligence (AI) solutions are available to assist banks in verifying business relationships for CB customers.


Global businesses are now identified through a North American Industry Classification System (NAICS) code, replacing the Standard Industrial Classification (SIC) system codes in 1997. NAICS codes are reviewed and updated every five years. The NAICS separates entities and businesses into industries based on relationships in their processes. Currently, there are 20 sectors and 1,057 industries.

Available AI solutions analyze NAICS codes for CB customers and compares the focal entity and counter parties of an alert to similar entities in transactional data that share common NAICS codes. The resulting AI analysis reports if the transaction makes sense from a valid economic purpose perspective, and if the entities are engaged in complementary lines of business.

In the graphic shown below, an AI agent was utilized to examine several months of CB transactional data after NAICS codes were matched to the entities. The AI agent reported expected and normal economic transactions which were graphically depicted by lines of varying widths based on statistics of prior similar transactions. Additionally, the AI agent reported the analysis in a table format with related transactional risk scores for each entity.