November 21, 2026, (Inside AI) — South Korean President Lee Jae Myung used a visit to a sprawling data centre site in the southeastern city of Ulsan on Saturday to deliver a blunt message about the country's artificial intelligence ambitions: the winner will be whoever moves fastest. The facility, under construction by SK Group in partnership with Amazon Web Services, carries an estimated price tag of 7 trillion won (about $5.2 billion) through 2029, according to South Korea's science ministry. Operations are scheduled to begin in the second half of 2027.
The presidential tour turned a routine infrastructure inspection into a statement of national strategy. Data centres sit at the base of the AI supply chain, driving demand for memory chips, networking gear and power systems. Lee framed the Ulsan project as a test case for whether South Korea can convert its semiconductor strength into broader AI competitiveness, and whether that growth can reach beyond the Seoul metropolitan area.
"The whole world is competing right now, and speed is vital," Lee said in a public video of his tour.
The scale of the build explains the urgency. SK Group Chairman Chey Tae-won said in September that the site's initial capacity would reach 103 megawatts by 2029, with plans to add nearly 900 megawatts in later phases. Chey also said the conglomerate is preparing additional partnerships with global technology companies as construction advances. The first phase alone would rank among the larger AI-focused data centre projects in Asia, and the expansion target would place it in a different tier entirely.
Read: South Korea Hits $1 Trillion in Exports on AI Chip Boom
Why Ulsan Matters Beyond Seoul
Lee's remarks carried a domestic political dimension. He noted that AI investment has clustered around Seoul and argued the Ulsan project demonstrates that regional growth is possible. South Korea has poured resources into AI research and chip manufacturing, but the benefits have concentrated in the capital region. A multibillion-dollar facility in a southeastern industrial city gives the government a visible counterexample.
The visit also surfaced a practical bottleneck. After being told that roughly 900 people were working at the site, Lee said advanced facilities were short of construction workers and that ministries needed to consider how to train the workforce for building data centres and AI infrastructure. That labor gap is not unique to South Korea. Data centre construction globally has run into shortages of electricians, welders and specialized tradespeople as hyperscalers race to build capacity.
The SK and AWS joint venture reflects a broader pattern in which Korean conglomerates pair with American hyperscalers to secure cloud capacity and AI compute. SK Group's affiliates include SK Hynix, one of the world's largest memory chip makers, which gives the conglomerate a direct stake in the hardware that fills such facilities. The partnership also gives AWS a local foothold in a market where demand for AI services is rising.
Lee's emphasis on speed lands amid a global buildout that has strained power grids and supply chains. Governments from the United States to the Gulf states have courted data centre investment with subsidies and fast-tracked permits. South Korea's approach ties the project to regional development and workforce policy, a combination that could become a template if the Ulsan site meets its 2027 target.
The president's visit did not include new financial commitments or policy announcements. Instead, it served as public pressure on ministries and the private sector to keep the schedule. The science ministry's cost estimate and the 2027 operations date now function as benchmarks against which the project will be judged.
For SK Group, the Ulsan site is part of a larger bet that AI infrastructure will absorb the output of its chip business and generate new revenue streams. For the government, it is a chance to show that AI investment can create jobs outside Seoul. Both goals depend on construction crews, power connections and equipment deliveries arriving on time.
Read: SMIC: AI Chip Supply Chain to Stay Tight Through 2027
Lee's message was directed as much at his own administration as at the companies involved. Training workers for data centre construction takes years, and the shortage he identified will not resolve quickly. The Ulsan project will test whether South Korea can pair its chip expertise with the physical infrastructure and skilled labor that AI computing requires.
If a research paper dropped this week, Mahesh has already read it. He covers AI tools and vibe coding for builders who need to know what is worth their time and what is not.