South Korea’s Ministry of Science and ICT said on October 5, 2026 that its sovereign AI foundation model contest will continue to a third round, while a new frontier AI development track and an industrial deployment track will be created. Officials proposed about 4.7 trillion won in mostly equity based public investment for the frontier program, focused on 10,000 Nvidia GPUs and training data, pending National Assembly approval.
This article aggregates reporting from 3 news sources. The TL;DR is AI-generated from original reporting. Race to AGI's analysis provides editorial context on implications for AGI development.
Korea is quietly building one of the most coherent state backed strategies for frontier AI outside the US and China. By confirming that the Sovereign AI Foundation Model Project will continue, while also proposing a separate 4.7 trillion won frontier program built around equity stakes and 10,000 high end GPUs, the government is signalling that it wants both competitive national models and a deeper industrial base around them.
The policy shift away from pure grants toward equity like structures is a big deal. It borrows from sovereign wealth and industrial policy playbooks, making the state a capital partner in frontier compute rather than just a buyer of research projects. That can reduce misaligned incentives and gives Korea more leverage in how resulting models are governed and deployed across domestic industries.
This also intensifies competitive pressure on local champions SK Telecom, LG AI Research and Upstage, which now have to navigate a landscape where GPU allocations, equity terms and deployment support may all be contingent on ministry decisions. For the broader race to AGI, Korea’s move adds another well resourced node pursuing sovereign scale models, increasing global compute demand and making coordination on safety and evaluations more complex.



