South Korea targets $380B AI infrastructure push led by Samsung, SK Group
In brief
- South Korea targets $380 billion in private-sector AI infrastructure investment
- Samsung and SK Group commit over 2 trillion won for AI and semiconductors
- AWS investing 7 trillion won in South Korean AI data centers through 2031
- Power grid bottlenecks threaten gigawatt-scale data center deployment
- Semiconductor cyclicality poses downside risk if AI demand slows
Samsung's Aggressive Expansion
Samsung Group plans to invest over 1,000 trillion won, roughly $648 billion, by 2026 and beyond. More than 350 trillion won of that total is earmarked specifically for AI infrastructure, including the construction of advanced data centers. The conglomerate's bet on AI is underscored by projections that Samsung Electronics will achieve an operating profit of 550 trillion won by 2027, driven significantly by surging demand for AI servers.
SK Group and Global Partnerships
SK Group is reportedly planning collective investments of up to 1,100 trillion won for semiconductors and AI data centers. Beyond domestic players, Amazon Web Services is committing an additional 7 trillion won, approximately $4.9 billion, specifically for AI data centers in South Korea from 2025 to 2031, bringing AWS's total investment commitment in the country to roughly 12.6 trillion won.
The scale of this buildout reflects a strategic reality: Samsung and SK Hynix together control a dominant share of the global high-bandwidth memory market, the exact type of memory critical for AI training and inference. Owning the memory supply chain while building data center infrastructure creates a vertically integrated advantage.
The Infrastructure Crunch
The ambition extends to gigawatt-scale AI data centers, facilities that would consume as much power as small cities. But there's a catch. Power infrastructure is the obvious bottleneck. South Korea's power grid wasn't designed with this kind of concentrated industrial demand in mind. Solving that constraint will require parallel investment in generation and transmission capacity—a slower, more regulated process than building chip fabs.
Cyclical Risk
The semiconductor industry has a long history of boom-and-bust cycles. The semiconductor industry has a long history of massive buildouts followed by painful corrections when demand doesn't materialize as expected. If AI adoption slows or consolidates around fewer players, South Korea's massive capital deployment could face the same reckoning. For now, though, the country's tech giants are betting heavily that the AI buildout is structural, not cyclical.
