“JIL’s take on AI”: Korea’s AI Leap

By John Ian Lau, Contributing Editor

In the high-stakes race for artificial intelligence dominance, South Korea is punching above its weight—or at least trying to. Seoul has recently unveiled a national AI and venture strategy, a blueprint that’s as comprehensive as it is aggressive. Korea has long been a model of execution, from time of electronics manufacturing and development: now in the AI race, it is blending sovereign compute infrastructure, capital infusions, global outreach, and regional decentralization into a plan aimed at spawning 10,000 AI and deep-tech startups and 50 unicorns or decacorns by 2030.

At its core, the strategy is operational, not just rhetorical. Korea has already committed roughly $949 million in initial spending on 10,000 Nvidia GPUs, including cutting-edge H200 and B200 models, as part of a push to amass 50,000 units by 2030 for a national AI platform accessible to startups, small businesses, universities, and researchers. Allocations start in February, with projects eligible for up to 256 H200s or 128 B200s, while 6,000 B200s are earmarked for developing indigenous foundation models. Nvidia’s parallel commitment, inked with local giants like Samsung, SK, Hyundai, and Naver, adds another 260,000 GPUs, fueling what Keay dubs “AI factories” with a combined investment nearing $3 billion.

Capital isn’t an afterthought. High-growth ventures can tap up to $68 million per company through the Next-Generation Unicorn Discovery project, alongside relaxed public procurement rules for B2G deals and milestone-based B2B collaborations. To globalize, Korea is planting startup campuses in tech hubs like Silicon Valley, Tokyo, Singapore, London, and New York, while fostering 10 provincial “startup cities” to dilute Seoul’s dominance. Pension funds are being channeled into venture capital, with a target of $27 billion in annual VC investment by reshaping the funding ecosystem. It’s a holistic pivot toward resilience, including a “Rechallenge Support Network” for failed founders— a cultural shift in a nation historically risk-averse. But the real benchmark is the AI superpowers: the US and China. Korea’s moves are bold, yet the scale pales in comparison, underscoring how mid-tier players must innovate smarter amid a duopoly that’s sucking up global talent, capital, and compute.

Take the US, where private AI investment hit $109.1 billion in 2024 alone—nearly 12 times China’s $9.3 billion and far outstripping Korea’s estimated $1.56 billion slice of global AI startup funding through Q3 2025. By 2025, AI captured half of all global venture funding at $202.3 billion, with the US dominating. Government muscle amplifies this: The Trump administration’s “America’s AI Action Plan,” released in July 2025, mandates AI adoption across 400+ federal agencies by September, potentially unlocking trillions in economic upside. A $500 billion infrastructure push, starting with $100 billion from partners like OpenAI, Oracle, and SoftBank for the “Stargate” project, dwarfs Korea’s GPU ambitions. Add $320 million from the Energy Department for AI in science and a $2.8 billion five-year ramp-up in federal AI R&D spending, and the US isn’t just competing—it’s defining the field. Unicorn count? The US minted at least 80 new tech unicorns in 2025, with over 160 AI-focused, led by behemoths like OpenAI at $500 billion valuation.

China, meanwhile, plays the long game with state-orchestrated precision. Beijing’s “New Generation AI Development Plan” eyes global leadership by 2030, backed by a $138 billion venture fund over 20 years for AI and quantum tech—averaging $6.9 billion annually. 15 11 In 2025, total AI investment reached $125 billion, comprising 38% of the global pie, with government contributing up to $56 billion and capex hitting $98 billion. A $47.5 billion semiconductor fund bolsters homegrown chips, sidestepping US export curbs. Unicorns: China added seven AI players in 2025, bringing its total to 70-75, though some estimates peg it lower at 19-36 globally, with China holding a significant share. ByteDance and emerging LLM firms like MiniMax highlight a narrowing tech gap, with ROI projections at 52% by 2030.

Korea, by contrast, boasts just 21-30 unicorns total in 2025, with AI standouts like Rebellions and FuriosaAI numbering in the single digits—though a new “K-NPU” project aims for five AI chip unicorns by 2030. Its $4.9 billion 2025 R&D budget and 8% AI funding hike for 2026 are commendable, but represent a fraction of the superpowers’ outlays. The AI market? Projected at $7.17 billion in 2025, en route to $53.87 billion by 2032—solid growth, but overshadowed by US enterprise spending alone at $37 billion on generative AI this year.

JIL’s Take: Korea’s strategy is a masterclass in targeted efficiency, leveraging partnerships and policy tweaks to maximize limited resources. But in a world where compute is king, Seoul’s 50,000-GPU goal feels quaint next to US hyperscalers or China’s state-backed blitz. For investors, this signals opportunity in Korea’s niche plays—like AI chips via Rebellions—but the real alpha remains in US innovation hubs and China’s scale. If geopolitical tensions escalate, Korea’s balanced approach could emerge as a hedge. Watch for execution: Ambition is cheap; delivery wins races.

Views are the author’s and not investment advice.