Minister Kim Outlines $1.6 Quadrillion AI Investment Plan

The National Economic Advisory Council held an open forum titled “Three Mega-Projects: Challenges for Becoming a National Growth Platform” on the 2nd at Ferrum Tower in central Seoul. In his keynote, Minister of Trade, Industry and Energy Kim Jung-kwan said that if 1,600 trillion won in private investment is properly deployed, it could break the low-growth trend and help South Korea lead the world in the AI era.

The planned private investment equals about 60% of gross domestic product (GDP), according to the article. That is higher than the GDP shares represented by construction of the Gyeongbu Expressway, at 2.4%, and the U.S. New Deal, at 40%. Kim said the large investments are taking place in regional areas and could help address three challenges together: low growth, regional imbalance and the AI revolution.

Lee Jun, head of the Strategic Industries Research Center at the Korea Institute for Industrial Economics and Trade, said semiconductor investment is tilted toward expanding memory production capacity. Of the 881 trillion won in private investment for semiconductors, he said, more needs to flow into the system semiconductor ecosystem. South Korea holds nearly 80% of the high-bandwidth memory (HBM) market share, but has a weak presence in non-memory semiconductors, which account for 76% of the global semiconductor market. Lee said the spread of custom HBM, chiplets (small chip components combined into a larger processor) and heterogeneous integration is blurring the boundary between memory and logic chips, creating an opportunity to extend Korea’s memory strength into system semiconductors. He added that AI data centers and physical AI are generating chip demand, and called for policy links to carry mega-project demand from fabless companies to back-end processing firms.

Changwon National University President Park Min-won said the gains from AI investment should accumulate in domestic manufacturing. He warned that domestic AI leadership could weaken if foreign big tech and robotics companies build factory data and work know-how into training assets. He proposed manufacturing data usage rights that set in advance where data is stored, how it may be retrained and reused, and the conditions for its return or deletion. Sogang University Professor Jung Byung-sun said the investment plans for fabrication plants and data centers lack the intelligence-building stage needed to turn them into productivity gains on factory floors. He reported that AI is used by 9.2% of South Korean companies with at least 50 employees, while 3.7% have embedded AI agents across their entire organizations.

Kim Seong-sik, vice chair of the National Economic Advisory Council, called the three mega-projects a major challenge to reshape the economy and industry. He said that with a semiconductor supercycle underway, South Korea should achieve a qualitative leap in its semiconductor and AI ecosystems and ignite its next industrial engine.

The National Economic Advisory Council held an open forum titled “Three Mega-Projects: Challenges for Becoming a National Growth Platform” on the 2nd at Ferrum Tower in central Seoul. In his keynote, Minister of Trade, Industry and Energy Kim Jung-kwan said that if 1,600 trillion won in private investment is properly deployed, it could break the low-growth trend and help South Korea lead the world in the AI era.

The planned private investment equals about 60% of gross domestic product (GDP), according to the article. That is higher than the GDP shares represented by construction of the Gyeongbu Expressway, at 2.4%, and the U.S. New Deal, at 40%. Kim said the large investments are taking place in regional areas and could help address three challenges together: low growth, regional imbalance and the AI revolution.

Lee Jun, head of the Strategic Industries Research Center at the Korea Institute for Industrial Economics and Trade, said semiconductor investment is tilted toward expanding memory production capacity. Of the 881 trillion won in private investment for semiconductors, he said, more needs to flow into the system semiconductor ecosystem. South Korea holds nearly 80% of the high-bandwidth memory (HBM) market share, but has a weak presence in non-memory semiconductors, which account for 76% of the global semiconductor market. Lee said the spread of custom HBM, chiplets (small chip components combined into a larger processor) and heterogeneous integration is blurring the boundary between memory and logic chips, creating an opportunity to extend Korea’s memory strength into system semiconductors. He added that AI data centers and physical AI are generating chip demand, and called for policy links to carry mega-project demand from fabless companies to back-end processing firms.

Changwon National University President Park Min-won said the gains from AI investment should accumulate in domestic manufacturing. He warned that domestic AI leadership could weaken if foreign big tech and robotics companies build factory data and work know-how into training assets. He proposed manufacturing data usage rights that set in advance where data is stored, how it may be retrained and reused, and the conditions for its return or deletion. Sogang University Professor Jung Byung-sun said the investment plans for fabrication plants and data centers lack the intelligence-building stage needed to turn them into productivity gains on factory floors. He reported that AI is used by 9.2% of South Korean companies with at least 50 employees, while 3.7% have embedded AI agents across their entire organizations.

Kim Seong-sik, vice chair of the National Economic Advisory Council, called the three mega-projects a major challenge to reshape the economy and industry. He said that with a semiconductor supercycle underway, South Korea should achieve a qualitative leap in its semiconductor and AI ecosystems and ignite its next industrial engine.