Korea's AI Push Shows Policy Gains, Private Investment Gaps
South Korea's government has made measurable progress in its first year pursuing top-three AI status, but the country still trails the United States and China by orders of magnitude in private sector investment.
President Lee Jae Myung's administration secured 260,000 advanced graphics processing units and tripled the national AI budget to 9.9 trillion won in 2026. It enacted the world's first comprehensive AI framework law and launched a foundation model project to develop domestic large language models.
The effort earned recognition. In April, Stanford University's Human-Centered AI institute named eight Korean AI models as notable work-the third-largest count globally. Artificial Analysis ranked Korea as the "clear #3 nation in AI" in January.
Yet private capital tells a different story. Korea attracted $178 million in private AI investment in 2025, ranking 12th worldwide. The U.S. recorded $285.9 billion-160 times larger. China invested $12.4 billion.
According to an OECD report from February, U.S. firms captured roughly 75 percent of global AI venture capital. Korea did not register separately in the tally.
Where Policy Support Falls Short
Industry officials say the government's support-GPU procurement, infrastructure spending, tax incentives-cannot offset the scale advantage held by American and Chinese tech giants.
"What matters most is whether enough capital can flow into the private sector," one AI industry official said. "It is difficult to gain a meaningful edge when Big Tech firms are driving the market with massive scale and financial power."
The foundation model project illustrates the gap between ambition and execution. The stated goal is creating an AI model every Korean can use. Achieving that requires reasoning capabilities matching OpenAI's ChatGPT, Anthropic's Claude, and Google's Gemini.
So far, officials say the project focuses on selecting leading companies without a clear plan for developing advanced reasoning capabilities at scale.
Leadership Gaps Emerge
Personnel instability has complicated implementation. Ha Jung-woo, the inaugural senior presidential secretary for AI policy, stepped down in late April to run in a parliamentary by-election. The position remains unfilled.
"Policy support is certainly welcome, but in the end, it all comes down to how to secure experts," another industry official said. "Finding someone capable of overseeing and leading the country's overall AI strategy is a major challenge."
The administration appointed former LG AI Research head Bae Kyung-hoon as Science Minister to oversee the strategy. Presidents have also met with OpenAI CEO Sam Altman, Nvidia CEO Jensen Huang, and other AI executives, securing investment pledges.
What Comes Next
Industry officials say the administration's policies are encouraging but need clearer goals, faster execution, and evidence of sustainability. The foundation model project will select two winners next year-a test of whether the approach can produce competitive results.
The real measure will be whether government support catalyzes private investment or remains a temporary subsidy. Korea has the technical talent and semiconductor manufacturing base. Whether that translates to AI leadership depends on capital flowing to startups and established firms at scale comparable to competitors.
For executives assessing Korea's AI trajectory, the gap between policy momentum and private sector investment remains the central question.
Learn more about AI for Executives & Strategy and Generative AI and LLM developments.
Your membership also unlocks:
