Who owns the AI boom? South Korea's generational reckoning
There is a Korean proverb that translates roughly as: even if you fill a house with gold, you still want silver. It captures, with disarming accuracy, the psychology of a nation that rebuilt itself from rubble into the world's twelfth-largest economy in a single generation — and which now, in 2026, finds itself gripped by a quieter anxiety. South Korea's technology sector is powerful beyond anything its founders imagined. Yet across the country, in university lecture halls, in the cramped apartments of Seoul's inner suburbs, and in the offices of a government racing to shape the next era of global production, a question has become impossible to avoid: who actually owns the AI boom?
The immediate answer sounds like a success story. Samsung, SK Hynix, and their satellite suppliers command central positions in the global semiconductor supply chain. Samsung alone operates the world's largest chipmaking complex at Giheung, south of Seoul. The country's semiconductor industry contributes an estimated 20 percent of GDP and accounts for more than 40 percent of total exports. Government projections suggest that AI-related investment across the private sector will surpass $40 billion by 2030. Seoul has pledged to train 100,000 AI specialists by the end of the decade and has committed billions in public R&D spending to support advanced chip development. On the face of it, South Korea is positioned to be not merely a participant in the AI economy but one of its principal architects.
And yet the anxiety persists. It surfaces in ways both granular and broad. In the granular: a graduate degree has become effectively mandatory for any professional job that pays enough to live on in Seoul. In the broad: a growing body of analysis inside South Korea suggests that the gains from AI-driven growth are accruing to a narrower segment of society than the country's post-war development model ever allowed. The generation that built South Korea — that worked in factories, studied through the night, mortgaged everything for their children's education — watched the rules of the game change. For them, AI felt like a reward. For their children, it increasingly feels like one more wall.
The generational arithmetic
Start with the numbers most South Korean economists treat as settled. Youth unemployment for those aged 15 to 29 has consistently run above the national average for the better part of a decade, even as overall employment figures stayed robust. The proportion of workers in non-regular employment — contract, part-time, or gig — has hovered near 40 percent. Housing prices in Seoul have reached levels that make ownership a distant aspiration for most under-40 households. The chaebols, the great family-controlled conglomerates that powered South Korea's industrialization, still dominate every strategic sector: chips, shipping, construction, finance. The difference now is that their dominance is increasingly legible as a structural barrier rather than an engine of opportunity.
South Korea's economic miracle operated on a specific compact: work hard, enter a major company, receive lifetime employment and housing benefits. That compact has frayed. The very firms whose capital expenditure headlines national statistics — Samsung's multi-year commitment to AI chip infrastructure, SK Hynix's memory expansion — are also the firms whose automation investments are reshaping the composition of the workforce. High-skill engineering roles proliferate in the announcements; the jobs lost to automated logistics, to AI-dispatched services, to algorithmic scheduling in retail, receive less fanfare. The employment projections accompanying South Korea's AI strategy do not provide granular breakdowns of who is expected to gain and who is expected to be displaced. The sources do not specify the government's own internal estimates for net job creation versus automation-driven displacement.
What is visible is the demographic signal. South Korea's total fertility rate — already the world's lowest — continued to decline through the 2020s. The government has offered financial incentives for childbearing and launched a $20-billion initiative to improve access to childcare. These policies are framed as responses to a general demographic crisis, but they are also, unmistakably, a response to the anxiety of a generation that does not believe the economy will offer them the same foundations their parents enjoyed. The AI boom has not reversed that anxiety. In many respects, it has sharpened it.
The political response
Seoul has not been passive. The administration of President Yoon Suk-yeol, and subsequently the policy priorities of his successor, have treated AI as a national security and economic question simultaneously. The 100,000-specialist target, the semiconductor R&D subsidies, the state-backed investment vehicles directing capital toward AI chip startups — these are not abstract ambitions. They are concrete instruments of a strategy that treats AI capacity as the equivalent of what steel production was to the 1960s: the foundational industry around which everything else either rises or falls.
The structural argument for South Korea's position is coherent. Memory semiconductors — DRAM and NAND flash — were the country's competitive advantage in the previous cycle. High-bandwidth memory, which stacks chips in ways that accelerate AI model training and inference, extends that advantage into the current moment. SK Hynix is a dominant global supplier. Samsung's memory division competes at the frontier. If the AI economy runs on specialized memory as much as on compute, South Korea is well-placed to shape the terms of that dependency.
But the structural argument carries a complication. The very firms best positioned to benefit from South Korea's AI push are the chaebols that have drawn sustained criticism for concentrating economic power. A strategy that channels public investment toward Samsung and SK Hynix reinforces existing structures rather than building new ones. The workers, the small and medium enterprises, the domestic service sector — the constituencies that successive governments have claimed to prioritise — do not feature prominently in the industrial logic of AI-led growth.
This is not a uniquely South Korean tension. Every advanced economy is navigating the same question: what does AI-driven productivity actually mean for the distribution of income and opportunity inside a society? But South Korea enters this conversation with a sharper inheritance problem than most. Its development model, celebrated globally for decades, produced extraordinary wealth alongside extraordinary concentration. The AI boom risks accelerating both.
The geopolitical dimension
The international environment adds further complexity. The United States has moved aggressively to rebuild domestic semiconductor capacity and has used export controls and investment restrictions to limit China's access to advanced chips and the equipment to make them. South Korea, as a treaty ally and as a host to U.S. military infrastructure, has been drawn into this architecture. Samsung and SK Hynix have faced pressure to restrict certain categories of chip exports to Chinese customers — a commercially painful constraint given that China is a major market for South Korean memory products. The geopolitical logic and the commercial logic are in tension, and the South Korean government has managed that tension by seeking exemptions, delaying implementation timelines, and presenting each accommodation as a temporary measure rather than a strategic commitment.
Taiwan's role in global chip supply remains central — TSMC fabricates the most advanced logic semiconductors — and South Korea's position as the primary supplier of the memory that sits alongside those logic chips in AI servers gives it a degree of structural leverage. But that leverage is conditional on the geopolitical stability of the Taiwan Strait. A conflict or a blockade that disrupted TSMC production would reverberate through every node of the global AI supply chain, South Korea included. South Korea's own border tensions — North Korean leader Kim Jong Un called on 18 May 2026 for the military to make the border with South Korea an "impregnable fortress" and to strengthen first-line units against the "sworn enemy", according to Reuters and French-language reporting by France 24 — are not the primary risk to AI supply chains, but they are a reminder that the country's strategic position rests on a security architecture that is under pressure across multiple dimensions simultaneously.
The China file complicates matters further. South Korea has significant economic exposure to Chinese demand — Samsung's memory division generates a substantial share of its revenue from Chinese customers — and a strategic relationship with the United States that carries security obligations. Navigating those two commitments has become one of the defining diplomatic challenges for Seoul. The current administration has broadly aligned with U.S. technology restrictions, but has also sought to preserve commercial space with Beijing. The result is a policy of calculated ambiguity: participating in export control frameworks while extracting maximum flexibility on implementation timelines and customer categories.
What the boom actually means
The question at the heart of South Korea's AI moment is not fundamentally about geopolitics or industrial strategy. It is about what a technologically advanced economy owes the people who live inside it.
South Korea has demonstrated repeatedly that it can build things that the world wants to buy. The question now is whether it can build an AI sector that generates broadly shared prosperity — or whether the crown will go to the firms and the senior executives who already sit atop the country's economic hierarchy, while the workers, the students, the families who sustain domestic demand are expected to absorb the displacement costs of a transition they did not design.
The country has time. Its industrial base, its institutional depth, its semiconductor expertise — these are genuine assets, not the product of marketing departments. If the AI economy rewards memory and systems integration as much as it rewards pure compute, South Korea enters the next decade in a structurally strong position. But strength and fairness are not the same thing. The South Korean development model produced one of the great economic transformations of the twentieth century. It also produced a society in which housing is unaffordable, in which children are too expensive to raise, in which the ladder that made the miracle possible has been pulled up. The AI boom does not automatically fix any of those things. At its worst, it could make them worse.
The Korean proverb insists on wanting more. That hunger built a powerhouse. Whether it can also build a future that feels worth wanting — for a generation that arrived after the miracle and is being asked to fund its costs — is the question South Korea has not yet answered.
—
This publication's wire coverage of Kim Jong Un's border directive and South Korea's AI ambitions appeared alongside parallel reporting from regional and specialist outlets. The Reuters framing foregrounded the defensive posture of North Korea's strategic communications; the Nikkei Asia reporting foregrounded domestic socioeconomic anxiety in unusually direct terms. Neither lens is wrong, but neither is sufficient on its own. The structural dimension — the concentration of AI gains inside a narrow corridor of chaebol-adjacent firms — is what neither wire treatment foregrounded, and that is where this story sits.
Wire provenance
This editorial synthesis draws on the following public wire/social posts:
- https://t.me/france24_fr/18488
- https://t.me/NikkeiAsia/19523
- https://t.me/france24_en