South Korea unveils plans for new AI data centers backed by ~$357.5B from SK Group, GS Group, and Naver, targeting 8.4GW initially and 18.4GW by 2035
President Lee Jae Myung speaks during a public briefing on the government's three flagship mega projects at Cheong Wa Dae in Seoul, Monday.
Context & Ripple Effects
South Korea’s AI infrastructure push has progressed from public AI and chip-support commitments to a growing pipeline of proposed data-center projects. Earlier coverage included a planned 3GW facility led by Voltai and a separate 250MW project involving Reflection AI and Shinsegae Group.
The new plan brings SK Group, GS Group, and Naver into a much larger, state-backed buildout. It also sits alongside long-term domestic semiconductor investment plans involving Samsung and SK Hynix, linking compute capacity more closely to the country’s chip-industrial strategy.
First-order effects
- SK Group, GS Group, and Naver become central private-sector backers of a national-scale AI data-center program, with an initial 8.4GW capacity target setting the immediate buildout agenda.
- The plan creates a far larger domestic target market for data-center construction, equipment, power connections, and AI compute services than the individual projects previously reported.
Second-order effects
- Existing and prospective Korean data-center projects, including the Voltai and Reflection AI-linked plans, will operate within a more competitive race for sites, grid capacity, construction capability, and enterprise AI workloads.
- The scale of planned compute capacity strengthens the strategic case for domestic chip makers and AI-service providers to align capacity, infrastructure, and customer offerings inside South Korea rather than treating them as separate investment tracks.
Third-order effects
- If execution matches the announced targets, South Korea could shift from backing discrete AI and semiconductor projects toward coordinating a national AI stack: chips, data centers, and locally controlled compute capacity.
- The constraint increasingly becomes infrastructure governance—especially whether power and grid buildout can keep pace with announced compute ambitions—rather than simply attracting AI capital.
The trend: This is part of the broader sovereign-AI trend in which governments and national champions treat large-scale compute infrastructure as a strategic industrial asset.