"South Korea Needs a Different AI Strategy from the US and China"... WEF's Assessment of Korea's AI

WEF Report:
"Limitations in Dominating the Entire AI Spectrum Like the US and China"
"Expanding Models and Services on Top of Hardware Strengths Is Crucial"

Image generated by Gemini

There is advice that, rather than attempting to dominate the entire spectrum of artificial intelligence (AI) as the United States or China does, South Korea-which has focused its investments on AI hardware-should adopt a more pragmatic approach. This would involve concentrating on its areas of strength and supplementing its shortcomings through international cooperation.

On January 27, the World Economic Forum (WEF), in a report titled "Redefining AI Sovereignty" jointly published with global consulting firm Bain & Company, described South Korea as "a country that initially concentrated investments in hardware sectors such as semiconductors and is now expanding efforts into foundation models and applications." The report also cited statistics showing that South Korea invested 2.2% to 3.3% of its gross domestic product (GDP) in the entire AI value chain from 2010 to 2024, evaluating the country as one that has built competitiveness based on its AI hardware and manufacturing capabilities.

South Korea's AI investment, amounting to 2.2% to 3.3% of its GDP, is lower than that of the United States (3.4% to 5.1%) and Singapore (3.1% to 4.6%), but higher than China (1.7% to 2.6%), which is also considered a leading AI powerhouse alongside the United States.

In the report, the WEF categorized the AI competitiveness strategies of countries worldwide into five types.

The United States and China were classified as "Global AI value chain leaders," making massive investments across the entire spectrum of AI infrastructure, models, and services. Singapore and the United Arab Emirates (UAE) were presented as representative examples of "Ecosystem builders," having established regional ecosystems through relatively balanced investments and policy coordination. Germany and France were described as "Selective players," focusing their competitiveness on certain segments of the AI value chain such as data and applications. India was evaluated as an "Adoption accelerator," prioritizing the expansion of app services and AI usage. Countries whose AI capabilities are still in the early stages were referred to as "Emerging collaborators."

Applying these categories to South Korea, the report suggests that the country is in a transitional phase-moving from a "Selective player" with strengths in specific domains toward becoming an "Ecosystem builder" by expanding investments in foundation models and applied services.

Based on these classifications, the report emphasized that "it is realistically impossible for most countries to directly own the entire AI value chain as the United States and China do." In fact, since 2010, more than half of global AI-related investments have been concentrated in infrastructure such as data centers, as well as apps and services, with AI infrastructure investments alone exceeding 600 billion dollars (approximately 864.18 trillion won). Currently, the United States and China together account for about 65% of total investments across the entire AI value chain.

Maria Basso, Head of AI Applications and Impact at the WEF Centre for AI Excellence and a contributor to the report, suggested that in this context, "AI sovereignty should be redefined not as the issue of direct ownership of everything, but as the ability to design and control AI in line with national values and norms, and to secure resilience that remains robust even in times of crisis." She particularly noted that, given the high barriers to entry for AI infrastructure and foundation models in terms of capital, energy, and supply chains, strategic interdependence based on international cooperation and partnerships could become the core pathway to competitiveness.