Creating an AI-Friendly Environment by Financial Authorities... Increased AI Utilization in the Banking Sector (Comprehensive)

To Activate AI... Guidelines and Testing Centers to Be Established

[Asia Economy Reporter Park Sun-mi] As financial authorities plan to create guidelines for the activation of artificial intelligence (AI) and establish a verification testbed, the channels for AI technology utilization in the banking sector, which has been preparing for digital transformation in finance, will expand further.

According to the financial sector on the 14th, preparations for AI guidelines in the financial field will begin within the first half of this year, and practical guidelines related to utilization by financial industry sectors will be created within the year. While big data infrastructure for AI activation will be established, a testbed for algorithm validity verification will also be created. Do Kyu-sang, Vice Chairman of the Financial Services Commission (FSC), announced this plan yesterday at the Digital Finance Council, stating, "Since the digitalization of finance is an irreversible change, policy efforts are needed to prepare for new risk factors."

According to the AI guideline research project results received by the FSC from Seoul National University Industry-Academic Cooperation Foundation, the guidelines are expected to be created with the goal of enhancing reliability throughout the entire process of AI system planning, design, development, evaluation, verification, introduction, operation, and monitoring related to AI system development, commercialization, and utilization, thereby contributing to AI activation.

Specifically, it was recommended that the guidelines include ▲establishing governance to assess and manage potential risks of AI systems ▲conducting social impact assessments of AI use cases and ensuring human controllability ▲reviewing the quality of training data at the AI system development stage ▲evaluating the legitimacy of personal information use and selecting models considering AI system explainability ▲assessing AI performance, fairness, non-discrimination, and explainability ▲supervising AI performance and evaluating safety and security.

Currently, AI technology utilization in the banking sector is at a nascent stage. Due to unclear legal regulations on AI use, there has been a strong call for the establishment of guidelines. Accordingly, once AI operation guidelines are introduced and a testbed capable of demonstrating AI technology is established, the path for AI utilization in the banking sector will open wider.

KB Kookmin Bank AI Experience Center

Active Exploration of AI Utilization Plans in the Banking Sector
Preparing to Introduce AI Bankers

Kookmin Bank is currently preparing to introduce AI bankers at its branches. Starting around July this year, it plans to sequentially introduce deep learning-based AI consultation kiosks at 4 to 5 bank branches. From November, the deployment will expand mainly to complex, large branches and unmanned branches. The AI kiosks, which will be installed next to ticket machines, aim to introduce, explain, recommend, and sell financial products tailored precisely to customers by reflecting their personal information. Once guidelines are established, it will be possible to go beyond relatively simple tasks such as window guidance and financial product information to advanced recommendation and sales.

Shinhan Bank is also accelerating the development of AI bankers. Through an agreement with Samsung Electronics, it is actively working on introducing the artificial human ‘Neon’ service and is exploring other forms of AI bankers. Earlier this year, it conducted a mid-year regular personnel reshuffle for 2,414 employees using its self-developed ‘AI Optimal Solution Algorithm’ technology.

Hana Bank operates the HiChatbot service, which enables chatbot consultations and product subscriptions through AI algorithms, the HiRobo service where AI algorithms manage assets instead of humans, and the HanaEZ overseas remittance service using big data technology and AI algorithms. Currently, efforts are underway to enhance utilization, and once AI guidelines are released, the scope of application in existing services can be further expanded.

Woori Bank established an AI division in July last year to research the application of AI technology in banking business and seek efficiency improvements in AI utilization. It is already operating a system to prevent financial fraud in advance and limit damage using AI technology. In July, it plans to build an AI-based market prediction system that will enable financial market trend analysis, portfolio recommendation, and rebalancing functions using AI. In the second half of the year, it plans to build an integrated AI consultation platform to advance AI consultation bots and chatbots using AI technology.

NH Nonghyup Bank completed the ‘Investment Product Incomplete Sales Inspection Enhancement Project’ through an AI technology-based Robotic Process Automation (RPA) system. It replaced manual inspection tasks with automated robots, securing accuracy and efficiency in detecting incomplete sales. AI consultation was introduced into the loan automatic extension system to automate the entire loan extension process, and services that monitor suspicious transactions using machine learning technology and allow additional verification at the Accident Prevention Center are also being prepared.