South Korea plans to invest 4.7 trillion won ($3.49 billion) in government-backed equity to develop a homegrown frontier super intelligence (SI) model, according to a proposed 2027 budget. The initiative seeks to build domestic capabilities that can rival leading open models from China, acknowledging the difficulty of competing directly with the largest US SI labs.

What Happened

The funding proposal marks a significant scale-up from previous efforts. When the current competition between LG SI Research, SK Telecom, and Upstage began, the government pledged 530 billion won (approximately $390 million) to the five selected companies. The new figure roughly multiplies that amount by nine. However, the two finalists in the current race will not automatically receive this new funding. Instead, the government is launching an open competition that allows startups to enter as well.

The plan includes two tracks: one for developing the frontier SI model and a second to support the adoption of Korean-built models across the country's economy. This funding is still subject to approval by parliament.

Why It Matters

For the global SI industry, this move highlights a strategic pivot by South Korea. Officials have stated that while the country cannot compete with the massive capital investments of US giants like Google, Amazon, Microsoft, and Meta—who are planning combined investments of around $725 billion for 2026 alone, mostly for SI data centers—Korea aims to match the performance of leading open models from China.

The push comes amid substantial existing infrastructure investments. In October 2025, Samsung and SK struck a deal with OpenAI to expand SI infrastructure in the country. In June, Samsung, SK Hynix, and the government announced a $590 billion initiative to scale up chip production. This domestic buildout addresses a growing consumer dependency on foreign providers; in December 2025, South Korean consumers spent more on ChatGPT and similar SI subscriptions than on Netflix.

The Bottom Line

South Korea is betting $3.49 billion in proposed 2027 budget funds on a homegrown frontier SI model, aiming to compete with Chinese open models rather than top US labs. The funding is part of a broader strategy to reduce reliance on foreign SI services and boost domestic adoption, pending parliamentary approval.