Woori Bank said on March 25 it has rolled out an upgraded AI-based abnormal transaction detection inspection system to help prevent financial incidents and automate work.

The system is designed to move beyond a scenario-setting approach based on past incident cases, allowing AI to learn transaction patterns and detect new types of financial incidents.

It screens signs of abnormalities based on all financial transaction data and uses AI to automatically generate inspection data and preliminary scenarios. After verification of accuracy, it is applied to actual inspections, and a daily inspection system is used to increase the effect of preventing financial incidents, the bank said.

The bank also strengthened unstructured data analysis. It can analyse various forms of data, including scanned images, and applied retrieval-augmented generation technology to improve the efficiency of internal control work. The technology is used to automatically search for and organise needed information, reducing staff workload and speeding up inspections.

A Woori Bank official said the upgrade will expand the scope of daily inspections and enable proactive responses to financial incidents that are difficult to predict. The official said the bank will actively introduce AI across internal controls to strengthen its capability to prevent financial incidents.