Banks bet on artificial intelligence services

A 38-year-old surnamed Kim used his lunch break to visit KEB Hana Bank recently for investment advice.

With help from an employee, Kim accessed the bank’s robot adviser, Cyber PB, through a tablet PC. He answered questions regarding his investment preferences and the purpose of his investment.

In seconds, Cyber PB had a portfolio recommendation, which Kim took to invest in a bond fund and overseas dividend fund.

The process took only 10 minutes.

KEB Hana Bank is planning to offer Cyber PC through mobile devices, which will no longer require investors like Kim to visit the bank itself.

Other financial institutions are also using artificial intelligence in their businesses.

Shinhan Card is planning to install a fraud detection system run by artificial intelligence as early as the end of the year.

The computer, installed with a deep learning technique, automatically detects anomalies in credit card spending.

For example, if a client’s credit cards are used at a convenience store overseas, the fraud detection system automatically analyzes the questionable patterns and cancels the payments.

The Korean financial industry has been competitively installing AI services, which gained wide recognition thanks to the Go competition between player Lee Se-dol and Google Deep Mind’s artificial-intelligence program AlphaGo.

Financial institutions are currently using AI services in credit-score evaluation, asset management and financial consultancy.

But developments of applying AI with deep learning technology, much like that of AlphaGo, are underway.

Industry insiders expect the access to financial services and the choices that investors can make will expand once AI services are widely used.

“There will be more choices in various financial products, including loans, while it will be easier to get asset management services,” said Kim Kyung-bin, an analyst with Korea Center for International Finance.

Shinhan Bank in July applied a self-developed credit-score model that was applied to its mobile bank service Sunny Bank.

The bank said the credit-score evaluation model, exclusive for mid-level interest rate borrowers, has applied a new analysis that incorporates the basic artificial intelligence technologies - big data and machine learning.

Through the AI analysis, the bank has been able to match loans for clients who would previously not be eligible.

Woori Bank is working on a chatting bot for employees that they can quickly reference financial products and services. The Industrial Bank of Korea (IBK) will be testing a financial consultancy bot for customers.

AI technology is also expected to expand to other financial industries. Mirae Asset Global Investment, for example, last week opened an AI research center at Korea University.

“The research center will be working on developing AIs that could be used in all spectrums, from research to asset distribution solutions to analyzing investors’ preference and even developing financial products and services,” said an Mirae Asset official.

A financial industry official, who requested anonymity, said companies are rushing to AI development because it not only helps in developing customized products but it also cuts costs.

The Korea Center for International Finance researcher Kim, however, warns that while the convenience and efficiency that come with AI will improve significantly, there is also the risk of attacks from hackers.

BY KIM KYUNG-JIN [lee.hojeong@joongang.co.kr]