South Korea pledges over $1 trillion for AI and semiconductor dominance

South Korean President Lee Jae Myung just unveiled an investment plan that makes every other government chip subsidy look like pocket change.

The total: more than $1 trillion. Over the next decade. Into AI infrastructure and semiconductor manufacturing on Korean soil.

For context, the US CHIPS Act committed $52 billion. The EU Chips Act allocated about $47 billion. South Korea is going 20 times bigger.

The numbers

The plan has three pillars. Samsung Group pledged 1,000 trillion won ($648 billion) over ten years. SK Group committed 550 trillion won ($356 billion) toward AI data centres and chip production. Combined, Samsung and SK are talking about 2,000 trillion won, roughly $1.3 trillion, according to the Straits Times.

Bloomberg reports the government is orchestrating at least $880 billion in private investment from companies. The full package including government spending exceeds $1 trillion.

Samsung Electronics and SK Hynix separately announced 800 trillion won ($518 billion) specifically for a new chipmaking cluster, per AP.

Why now

Lee framed the push as a race against time. “We must secure our nation’s domination in the AI boom,” he said, according to Al Jazeera. The urgency is real. SK Hynix currently dominates the global market for high-bandwidth memory (HBM), the chips that make AI accelerators like NVIDIA’s GPUs work. Samsung is the world’s largest memory chip maker overall. If either loses ground to competitors in Taiwan, the US, or China, Korea’s entire economy takes the hit.

The timing also reflects market momentum. South Korea’s stock market is up roughly 200% year over year, fueled by AI demand. Samsung and SK Hynix, the two most valuable Korean stocks, have each rallied more than 186%.

The political catch

The plan includes a new semiconductor manufacturing hub in the Honam region, the rural southwest. That happens to be Lee’s political stronghold, where 85% of voters backed him in the last presidential election.

The opposition is not happy. They accuse Lee of choosing the location for political reasons rather than industrial logic. They claim the government is pressuring memory chipmakers to build in a region that may not be the most commercially viable.

Lee defended the choice over a series of posts on X over the weekend. His argument: the southwest has rich, untapped power resources that the chipmaking industry desperately needs. Data centres and fabs are power hogs. If the region has cheap electricity and available land, the industrial logic follows.

Both things can be true. It can be a politically convenient location and still make engineering sense. The question is whether chipmakers would have chosen it without government pressure.

The global AI arms race

This announcement fits a pattern. Every major economy is scrambling to build domestic chip production and AI compute capacity. The US, EU, Japan, India, and now South Korea have all launched massive semiconductor subsidy programs since 2023.

But the scale here is different. South Korea’s GDP is about $1.7 trillion. Pledging $1 trillion over ten years means committing roughly 6% of annual economic output to AI and chips. That is not a subsidy program. It is a national survival strategy.

Samsung and SK have the most to lose if AI demand shifts. Samsung is fighting to catch up to SK Hynix in HBM, the most profitable segment of the memory market right now. SK Hynix is planning a Nasdaq ADR listing that analysts think could value the company near $1 trillion.

What could go wrong

Ten-year investment pledges are political documents, not contracts. Samsung announced a similar 450 trillion won plan in November. Companies routinely promise large numbers to secure government goodwill, then scale back when market conditions change.

The bigger risk is execution. South Korea faces the same bottleneck everyone does: not enough skilled engineers, not enough power, and not enough time. You can pledge all the money in the world. Building a fab takes four years minimum. Training the workforce takes longer.

Lee’s bet is that the money and the political will can compress those timelines. The market seems to believe him, at least for now.