Bank of Kyoto Puts a DataRobot Anti-Money-Laundering Detection Model and Investigator Dashboard Into Operation

Bank of Kyoto, a regional bank headquartered in Kyoto, began full operation in July 2026 of an AI-based system for detecting suspicious transactions as part of its anti-money-laundering work, NTT Data announced on 14 July 2026. NTT Data supported the bank in building the AI model and a business intelligence dashboard around it. The model draws on transaction data, customer data and records of past suspicious transactions that the bank has accumulated on Service Innovation Core (SIC), a data utilisation platform from NTT Data that is also used by several other Japanese regional banks. Using DataRobot, the bank trained a model specific to its own business on those historical cases. Bank staff led the model-building work and fed their operational know-how into it, while data scientists from NTT Data's AI Success programme supported them, so that the bank can keep improving the model itself. The second part is a dashboard that links the model's predictions with the data on SIC and shows, in one place, the customer information, transaction details and other material investigators need to decide whether a transaction is suspicious. Before, that information was spread across several systems and had to be gathered and checked by hand after each alert. In a validation exercise that began in December 2025, the system detected suspicious transactions that the bank's conventional methods had not caught, and the work of collecting and checking information after a detection became about 30% more efficient. NTT Data says it will continue to combine SIC, DataRobot and its AI Success support to help other financial institutions use their data in the same way.