South Korean President Lee Jae Myung committed 4.7 trillion won to developing a sovereign AI model he described as targeting the "world's highest performance," part of the country's largest-ever budget presented to the National Assembly on Monday1.
The sovereign-model allocation sits inside a broader 21.3 trillion won package for semiconductors, physical AI, and AI data centers — nearly double the 2026 level2. Total government spending for 2027 is set to rise 12.8% to 820.9 trillion won ($610.14 billion).
Lee pledged to "boldly invest in AI and future growth engines". The $15.97 billion earmarked for AI, chips, physical AI, and data centers in the 2027 budget is nearly twice the 2026 amount. Alongside the hardware and model investments, the government plans to train 2,000 specialists to lead semiconductor design and regional chip industries.
Lee framed the budget against what he said was South Korea's approach toward $1 trillion in annual exports, driven by strong semiconductor and broader industrial performance.
The announcement lands as U.S. data center politics intensify domestically. Candidates backing data center moratoriums appeared competitive in up to 15 of 36 U.S. governors' races heading into Tuesday's elections[1].
ANALYSIS South Korea's near-doubling of its AI and semiconductor budget in a single fiscal year converts an export-led chip advantage — anchored by Samsung Electronics and SK Hynix — into a state-backed compute and model-development strategy, linking industrial policy to sovereign AI capability.