Artificial Intelligence Transformation (AX) is one of the most frequently mentioned keywords in industry today. The Korean government is also accelerating its goal of becoming a top‑three global AI powerhouse by placing sovereign and vertical AI strategies at the forefront. However, actual cases where AX leads to real competitiveness remain rare. This is because the method of Digital Transformation (DX), which underpins AX, still remains in the past.

DX in Korea is still driven mainly by in‑house development or on‑premise system construction. It takes one to two years to plan, develop, build, and operate a project. The problem is that this approach struggles to keep up with rapid technological changes. During the course of a project, both the technology landscape and market expectations continuously evolve. It is practically impossible to predict all of these changes in advance and reflect them in design. Most DX projects end after meeting the requirements of a specific point in time, and as new changes arise, another project must begin again.

This structure also reveals limitations in terms of data. Data generated in in‑house and on‑premise environments is limited in scale and density, and because it remains inside each company, it does not accumulate into industry‑level vertical big data. In such circumstances, even if AI is introduced, the actual utility often falls short of expectations. In other words, true AX is difficult to achieve if AI is adopted without first changing the underlying DX structure.

The solution lies in shifting the direction of DX. DX should not be treated as a one‑time project, but rather viewed from the perspective of platforms and Software‑as‑a‑Service (SaaS). When solutions are delivered on platforms in SaaS form, DX can evolve through repeated accumulation and expansion. It quickly reflects user needs, modifies structures, and reuses these improvements to generate further innovation. Furthermore, when services are used beyond the boundaries of a single company or country, highly dense vertical big data begins to form. The DX‑based ecosystem naturally expands into AX.

Nvidia CEO Jensen Huang also suggested during the CES 2026 keynote that the software (SW) environment will fundamentally change. Whereas people once programmed software, the structure will shift toward building, training, and running software on AI systems. This does not merely mean that AI will replace coding. Rather, it means the platforms that host software will themselves operate on AI, allowing the structure and functionality of software to be continuously reconfigured.

This trend is already well‑established in global markets. Global companies such as SAP, Salesforce, and ServiceNow have combined AI and no‑code environments with their platforms to create thousands of applications. In contrast, DX in Korea still tends to remain inside individual companies. The gap stems not from technology itself but from differences in structure.

If AX is promoted without sufficient DX foundations, it is difficult to achieve the expected level of impact. Conversely, DX continuously accumulated on top of platforms expands its usefulness over time and becomes a solid foundation for AX. The key to becoming an AI powerhouse lies not in simple technology adoption, but in the structure of innovation. This is why Korea needs a national AI leader that builds a new ecosystem using platforms and SaaS as its core strengths.

Lim Hyun‑gil, CEO of STEG Inc.