Capital Research Institute: "Financial AI Guidelines Must Increase Specificity and Timeliness"

As interest in the use of artificial intelligence (AI) in the financial industry grows, there have been suggestions to further specify and enhance the timeliness of financial AI guidelines.

On the 26th, Seongho Noh, a research fellow at the Korea Capital Market Institute, stated during a keynote presentation at the 'Utilization of Generative AI in the Financial Sector' seminar hosted by the Capital Market Institute at the Yeouido Financial Investment Center, "Domestic financial AI guidelines emphasize stability, showing consistency with cases from major overseas countries. Moving forward, it is necessary to further specify and enhance the timeliness of these guidelines to balance the stable use of AI and innovation in the financial industry."

Researcher Noh explained, "In Korea, since the publication of the 'AI Operation Guidelines in the Financial Sector' by the Financial Services Commission in 2021, guidelines for the safe use of AI-based services have been presented at multiple levels," adding, "It is characterized by specifying the relationship between internal controls of financial institutions and AI governance, as well as requiring periodic data quality verification and improvement."

He also pointed out, "Most financial AI guidelines have unclear standards for violations or, even if standards exist, the methods and scope of sanctions for violations are not clearly specified. In particular, follow-up measures are insufficient when financial institutions develop or have already applied AI services that violate the guidelines, such as withdrawing the services and compensating for damages caused."

Accordingly, there is a call for the process of specifying industry-specific regulatory and supervisory standards within the guidelines. Researcher Noh stated, "Some guideline principles focus heavily on comprehensive stability assurance, which restricts the introduction of innovative services. Abstract principles require significant time and effort for interpretation at the practical level, and due to the broad definition of stability, considerable costs are expected from the development stage of new innovative services."

Furthermore, he noted, "The guidelines rely on the level of technology development known at the time of their enactment, which risks low timeliness and may fail to respond to rapidly changing AI technology trends. Another issue is the existence of many overlapping areas between supervisory regulations applied to detailed sectors of the financial industry and newly announced guidelines, necessitating a review from the perspective of their relation to basic laws that have been enacted or are in the drafting stage."

On the same day, Jeonghoon Lee, team leader of the Cloud Business Division at Koscom, who gave a presentation on 'Demonstration Cases and Development Directions of AI Services in the Financial Industry,' explained the current status of AI service innovation finance designations through the financial regulatory sandbox and introduced AI-integrated services that Koscom is promoting.

This content was produced with the assistance of AI translation services.