South Korea Examines Agentic AI's Impact on Software, Jobs, and Regulation

South Korea’s Ministry of Science and ICT convened an expert roundtable on the era of agentic AI and its implications for the software industry and talent development. Officials and industry experts assessed the immediate shock to the sector and the longer-term implications for pricing, accountability, and workforce strategy.

Experts at the session said agentic AI can dramatically accelerate software development, shrinking projects that once took three years to about 40 days. The discussion cited the January release of Anthropic’s agentic AI "Claude CoWork" as part of a broader market shift, noting a roughly $1 trillion drop in the market capitalization of U.S. software firms and fueling talk of a so-called SaaS apocalypse.

Agentic AI combines the capabilities of generative AI with planning, decision-making, tool use, and execution, enabling it to perform tasks with limited human instruction. At the briefing, participants noted a real-world move toward AI-driven production and the possibility of a phase in which software costs approach zero as automation scales.

As AI-driven production expands, the mix of work in software development is changing. Field observations suggest the number of people working directly with a product owner is declining, and gaps between firms and individuals are widening based on how adept they are at deploying AI agents and related tooling. The industry is shifting from traditional software to AI-enabled software and services.

Policy questions dominated the discussion: how software projects should be priced when AI contributions are significant, who should own the final deliverable, and which new areas will demand government support as priorities. Officials said the government must adapt regulatory and institutional frameworks to reflect AI-enabled production realities.

Talent development was another major focus. Experts argued for transforming university curricula away from coding-centric training toward design and verification competencies—teaching problem definition, system design, and validation capabilities. They also called for closer industry–public data access, stronger infrastructure support, and expanded academia–industry collaboration to prepare the workforce for AI-augmented software.

The Ministry announced the formation of an advisory panel to establish AI ethics principles designed to promote human-centered and responsible development and use of AI. The plan is to draft an ethics framework by mid-April and finalize it by June after gathering input from the public, businesses, and other government entities.

For U.S. readers, these developments matter beyond Korea because they signal a global shift in software development, productivity, and talent needs driven by AI. The rapid acceleration of AI-enabled software could affect international supply chains, competitiveness of tech firms, and the regulatory landscape for cross-border collaboration on AI safety, ethics, and standards. The evolving definitions of responsibility and ownership in AI-produced software are likely to shape markets, investment, and policy coordination across the United States and Asia.