To achieve this, we analyzed the current status of domestic and international
classification systems, designed a multi-layered classification system centered on the
software, service, and hardware structures of the AI industry, and verified its
applicability to actual industry statistics and policies.
The study found that the AI industry cannot be adequately explained solely through a
technology-centric approach and requires an ecosystem-based definition encompassing
the entire value chain spanning AI software, service, and hardware. Accordingly, this
study divided the AI industry into three levels and subdivided industrial activities into
each sub-sector, establishing a foundation for understanding the current state of the
industry and generating statistics. Furthermore, by considering the potential for
integration with the existing Korean Standard Industrial Classification (KSIC) system,
we propose an improvement plan, ensuring both policy effectiveness and statistical
usability. The findings of this study can serve as supporting data for various policy
areas, including the government's AI industry promotion policy, digital transformation
strategy, R&D investment plans, and human resource supply and demand policies.
Specifically, by clarifying the scope of the AI industry and enhancing the reliability of
industry-specific statistics, the accuracy of policy goal setting and effectiveness
analysis can be enhanced. Furthermore, the classification system proposed in this
study offers a potential foundation for developing into a national industrial statistics
standard through collaboration among relevant organizations, including Statistics Korea,
the Ministry of Science and ICT, and the Ministry of Trade, Industry and Energy.
Future tasks include: first, empirical application and data-based verification of the
improvements proposed in this study. Through the participation of various
stakeholders, including industry, academia, and government agencies, detailed criteria
for classification items should be specified and their suitability to actual industrial
activities should be reviewed. Second, to respond to the rapid evolution of AI
technology and the proliferation of industrial convergence, a system for periodic
revision and update of the classification system is required. As the AI industry
continues to expand with the emergence of new technological paradigms such as
generative AI, autonomous intelligent systems, and edge AI, a flexible maintenance
and supplementation system is necessary. Finally, future research requires a detailed
design of a statistical production system linked to the AI industry classification system
and an in-depth review of international standardization measures. By linking with
discussions in international organizations such as the OECD, EU, and ISO, Korea's AI