Seven firms pledge $2 billion for Korea chip and AI projects

By Grace Kim

Seven global companies have pledged a record $2 billion for Korean semiconductor, physical AI, data center and energy projects in 2027, with 97% of the funds directed outside the greater Seoul area.

The $2 billion is the largest investment commitment secured at an annual forum hosted by South Korea's Trade Ministry and the Korea Trade-Investment Promotion Agency. Announcements came Wednesday at the 22nd investment summit in southern Seoul. The three-day event runs through Friday at the Grand InterContinental Seoul Parnas hotel.

The allocation forms a 97-to-3 split between nonmetropolitan regions and greater Seoul. It supports the government's push to spread AI-driven industrial growth beyond the capital.

“Korea has two national priorities: putting AI to work in manufacturing and helping regions across the country grow through AI transformation in manufacturing,” said Kim Tae-hyung, commissioner of Invest Korea, in Wednesday's opening remarks.

Kim said AI should help factory workers catch problems earlier. It should adjust processes before materials are wasted and keep production moving.

What did the seven companies pledge?

Nscale, Air Liquide and Amkor Technology lead the $2 billion package. Much of it targets industrial gases, materials and equipment used in advanced packaging for AI chips, including high bandwidth memory (HBM).

• Nscale, a British AI infrastructure company, plans to build an AI data center in Korea.
• Air Liquide, a French industrial gas supplier, plans to produce specialty gases used in chipmaking.
• Amkor Technology, a U.S.-based semiconductor packaging and testing company, will expand its advanced packaging facility in Gwangju.
• Germany's Henkel will establish a research and development center for advanced packaging materials used in HBM. It will also work with Korean chipmakers.
• Japan's Tokyo Electron will invest in a semiconductor equipment and materials plant.

These projects support Korea's ambition to use its manufacturing base as a testing ground for industrial AI. Foreign capital would extend beyond chip production into data centers, energy facilities and research capabilities needed for AI expansion.

What supports the AI infrastructure?

The data-center push also requires energy and research infrastructure, extending the announced industrial buildout beyond semiconductor factories and into the power systems needed for large AI deployments.

• Singapore-based Gurin Energy and Norway's Deep Wind Offshore plan to develop renewable energy complexes.
• Edwards, Valeo and Ibiden plan to strengthen their research and development presence in Korea. Their areas include semiconductors, future mobility and advanced materials, alongside expanded technology partnerships with Korean companies.

What does the $1.2 trillion figure cover?

Trade Minister Park Jung-sung said: “Companies have announced regional investment plans totaling more than $1.2 trillion, including three major projects in semiconductors, physical AI and AI data centers.”

The $1.2 trillion figure covers companies' announced regional investment plans. The $2 billion represents the commitment secured at this summit.

Park expects manufacturing facilities across the country to become AI-powered factories. He said industrial complexes will evolve into innovation hubs. He hopes those changes will create opportunities for global companies to test and commercialize their technologies.

The disclosed numbers represent investment commitments and plans. They do not establish operating capacity or revenue. The announcement specifies neither process nodes, wafer sizes nor output levels.

Foreign-invested companies accounted for 21 percent of Korea's exports and 6 percent of employment in 2025, according to government data. The new projects would extend that footprint into upstream inputs and enabling infrastructure.

What will determine the outcome?

Execution will determine whether the 2027 commitments add measurable HBM packaging capacity, expand local production of specialty gases and chip equipment, and broaden the supplier base supporting AI-driven manufacturing.