Sony Bank to use AI to bolster lending operations
AI is set to help the bank shorten the loan screening period from 2-6 days to as little as 60 minutes.
As artificial intelligence (AI) solutions take over mundane and repetitive tasks at financial services companies, the number of banks that start applying such technologies in their lending departments is growing. The latest firm to join this trend is Sony Bank Inc., a wholly owned subsidiary of Sony Financial Holdings Inc (TYO:8729).
Sony Bank is set to start using an AI application which was originally developed by the bank itself, to automate preliminary screenings for mortgage loans from May 10, 2018.
By using AI to learn from loads of previous data and reading in and making adjustments for new data, Sony Bank has succeeded in automating screening to a level of precision approaching that of human screeners. The new approach aims to accelerate the screening process, and shorten preliminary screening period from the current around two-six days down to as little as 60 minutes. By using AI to make credit decisions automatically that previously relied on humans, the bank hopes to enhance operating efficiency.
The new technology employs the ensemble method, a machine learning technique creating multiple models and combining them to produce improved results. As a result, the model offers a higher level of predictive precision.
Other Japanese banks have also tapped the AI capabilities in order to improve their lending services. For instance, in September last year, J.Score, a joint venture of SoftBank Group Corp (TYO:9984) and Mizuho Bank, said it would use AI to offer loans tailored to the specific needs and life style of each consumer. The service will decide on consumer loans based on an AI score. In particular, an AI technology is responsible for the calculation and analysis of the score.
And, in April this year, Shizuoka Bank Ltd (TYO:8355) announced that it will conduct an experiment using AI technology in an effort to enhance its corporate lending services.
The experiment will run until June 30, 2018 and will use the “SX Score”, developed by SecondXight Inc, an expert in financial data analysis. The bank will analyze various data, including information about financial transactions. The “SX Score” is a credit risk assessment model, which takes into account certain characteristics of the clients and partners of the bank. Based on the AI technology, the bank will be able to better target its services and to consider more carefully the size and types of loans it offers to its clients, such as medium and small businesses.