South Korea Bets 200 Trillion Won on AI and Megaprojects

Seoul’s Semiconductor Windfall Has a New Address

South Korea is about to do something it has never done at this scale: park a massive tax surplus in a government-controlled investment fund instead of using it for general spending or debt repayment. The Future Response Fund, set at roughly 200 trillion won ($135 billion), will channel windfall revenue from the current semiconductor supercycle into strategic bets on AI, next-generation energy, and regional development.

The logic is straightforward — and unusually aggressive for a country that has historically relied on export-led growth and market signals rather than explicit industrial policy. When tax revenue runs above trend, especially from a sector like semiconductors that booms and busts with brutal regularity, the money gets saved. Then it gets deployed when the next crisis hits — or, in this case, when the next opportunity appears.

That dual purpose is the fund’s defining feature. It is both a growth engine and a fiscal shock absorber. If domestic tax collections fall below the baseline trajectory, the fund can transfer money back into the general account. If they exceed it — as they likely will if the chip cycle stays hot — the surplus flows into investment projects.

The Math Behind the Money

The baseline calculation comes from the Ministry of Economy and Finance. Based on 2025 domestic tax revenue of 330.2 trillion won and applying the 10-year average annual growth rate of 6.1 percent twice, the trend benchmark for 2027 domestic taxes lands at 371.7 trillion won. Projected national tax revenue sits at 495 to 522.5 trillion won under a 90–95 percent domestic share. That creates a surplus of roughly 123 to 151 trillion won — before adding this year’s estimated excess revenue of around 50 trillion won.

Add in leftover world surpluses and restructuring savings from education grants, and the total approaches 200 trillion won.

But here’s what most international observers miss: the fund is designed to grow over time. If national tax revenue climbs toward 600 trillion won, the surplus window opens to 218–248 trillion won. The government plans to continuously funnel above-trend and excess revenue into the fund each year, meaning it could become a permanent, compounding feature of Korea’s fiscal architecture rather than a one-off intervention.

That makes it Korea’s first true sovereign wealth fund — not from oil or commodity exports, but from technology-cycle taxation.

The Three Mega Projects and the AI Gamble

The fund divides into four operational accounts, but the growth momentum account will absorb the lion’s share. Its mandate is explicit: push Korea into the global AI top three.

That means investing in frontier AI research, physical AI — robots and automation systems that interact with the real world — and AI data centers (AIDCs), which require enormous power and water infrastructure. The government also earmarks funding for seven so-called SEED technologies: small modular reactors (SMRs), nuclear fusion, space and aerospace, quantum computing, advanced biotech, and two unspecified categories.

The power and water requirements for AIDCs are the unglamorous bottleneck that everyone talks about but few plan for. An AI training run at the scale of Korea’s ambitions will consume more electricity than several mid-sized cities. The government acknowledges this by directing fund capital toward the physical infrastructure behind the technology — not just the models themselves.

For context, South Korea currently ranks outside the top three in global AI capacity. The US dominates. China is closing the gap fast. Japan has its own ambitions. A 200 trillion won commitment to close that gap represents one of the most concentrated industrial policy bets in East Asia this decade.

Who Wins and Who Loses

The winners are clear: AI startups that can secure government backing, semiconductor supply chain companies that serve AIDC construction, regional development zones that attract new infrastructure, and young Koreans facing historically high unemployment who gain subsidized housing and job training. The youth account specifically targets employment, housing, and marriage-childcare support — a direct shot at Korea’s 0.6 fertility rate, which is collapsing regardless of whether the government tries to fund its way out of it.

The losers are harder to name precisely but visible in structure. Taxpayers who would prefer the surplus used for debt reduction face a fund with loose congressional oversight — spending plans can be adjusted without legislative approval within a 20 percent range. Critics are already calling it a fiscal slush fund. Debt-to-GDP advocates argue the money should go toward paying down the 55 trillion won in projected interest costs that Korea faces as rates remain elevated.

The government’s defense is pragmatic: waiting for perfect fiscal conditions means missing strategic windows. US CHIPS Act funding and Chinese state-directed AI investment don’t respect Korea’s budget calendar.

The Unresolved Question

The biggest uncertainty isn’t the fund’s size or even its political feasibility — both are already baked into the 2027 budget plan. It is whether a government-directed investment vehicle can actually pick winners in AI and deep tech.

Korea has tried this before. The Chaebol-era industrial policy of the 1970s and 80s produced Samsung and Hyundai but also many state-backed failures that required bailouts. The 1997 Asian financial crisis exposed the cost of that approach. What makes this cycle different is that the surplus comes from a genuinely structural shift — semiconductor demand driven by AI — rather than temporary commodity prices. That may give the fund more durable footing.

But the fund’s success still depends on execution. Will the youth account actually move the needle on hiring and housing? Can the government direct capital toward AI startups without recreating the crony dynamics of older industrial policy? Will the loose oversight provisions lead to waste, or will they enable the speed Korea says it needs against US and Chinese competitors?

The answer will matter far beyond Korea. A successful pivot toward AI and deep tech, funded by a sovereign-style surplus mechanism, could become a model for other middle-income countries trapped between US and Chinese technological dominance. A failure would reinforce the view that Korea’s growth engine is stalling and its policy toolbox is empty.

The fund’s first disbursement decisions should come within weeks as the 2027 budget is finalized. Watch those allocations closely.