National AI Strategy Committee: "Success of Korean AI Depends on High-Quality Data"

Meeting Held with Leading National AI Companies

Moonyoung Lim, Vice Chairman of the National Artificial Intelligence Strategy Committee, Yonhap News Agency

The National AI Strategy Committee announced on April 2 that it held a meeting with leading national AI companies to discuss strategies for securing high-quality data.

This meeting was organized to listen to challenges related to data acquisition, which is one of the key factors for the success of the ‘Korean Proprietary AI Foundation Model (Dokpamo) Development Project’ involving leading national AI companies. Attendees included LG AI Research, Upstage, SK Telecom, and Motif Technologies.

LG AI Research emphasized the need for reasonable regulatory reforms to expand data utilization, which is core to AI competitiveness. The company stated that, with the government’s future efforts to improve various systems including copyright, it will accelerate the development of a globally competitive Dokpamo through public-private collaboration.

Upstage highlighted the urgency of building a data infrastructure to strengthen Korean-language AI competitiveness. The company called for the establishment of clear regulations for the use of works for AI learning purposes—such as the digitization of legal deposit books at the National Library of Korea and a text and data mining (TDM) exemption provision. It also stressed the need to build datasets focused on expert reasoning processes, such as solutions and explanations for national qualification exams.

SK Telecom proposed improvements regarding the overall process of data acquisition and utilization. The company suggested that the verification and refinement system for shared data provided by the government should be advanced so that the data can be applied directly to training without further reprocessing. It also recommended securing domain-specific data, expanding authoritative Korean evaluation datasets for objective model performance verification, resolving licensing issues to promote the use of publicly produced data, and supporting the reduction of burdens on the private sector in handling sensitive information.

Motif Technologies pointed out that the government’s data support projects are currently focused on securing raw data, and requested expanded support for post-processing and refinement of data, which are essential for actual AI model training.

Moonyoung Lim, Standing Vice Chairman, stated, “The success of the Dokpamo project, with the participation of leading national AI companies, marks the starting point for achieving AI for all and an AI-based society. The issue of securing data is not just the responsibility of a particular ministry or agency, but a national challenge. The committee will take the lead in mobilizing capabilities and providing effective, timely support to address this.”

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