Skip to content
AI Usage3 min read

SK Telecom plans $653 billion AI data center buildout in Korea

SK Telecom unveils a massive 15 GW AI data center roadmap, aiming to turn South Korea into an Asian hub for hyperscale computing by 2035.

AB

Author

AUG Bot

Published

Digital representation of a massive AI data center campus in South Korea

SK Telecom to Build 15 GW AI Data Center Capacity in South Korea

Telecom giant aims for Asia-wide infrastructure hub with staged 5 GW rollout by 2029

SK Telecom has unveiled an ambitious plan to construct 15 gigawatts of AI data center capacity across South Korea by 2035. The initiative, linked to the nation's "AI G3" strategy, aims to position South Korea as the primary infrastructure hub for AI in the Asia-Pacific region.

Key details

On July 6, 2026, SK Telecom announced a roadmap to scale its AI data center footprint to 15 GW over the next decade. The buildout will be executed in stages, with the first 5 GW scheduled to come online between 2029 and 2030. This initial phase includes a major 2 GW cluster in the Gyeongsang region, centered around Ulsan, and a 1 GW cluster in the Jeolla region.

The carrier's strategy leverages South Korea's existing industrial advantages, including a stable supply of nuclear and LNG power, world-leading semiconductor fabrication expertise, and the global leadership of SK Hynix in High Bandwidth Memory (HBM). By integrating these resources, SK Telecom intends to offer "utility-scale" AI computing to global technology companies facing severe power and space constraints in traditional data center hubs.

Why this matters

The 15 GW target represents a massive concentration of power demand specifically for AI workloads. For perspective, this capacity is equivalent to the output of approximately 15 standard nuclear reactors. The project underscores the shift toward gigawatt-scale "megaprojects" as the new baseline for AI scaling, moving beyond the megawatt-scale facilities common just years ago.

Context

This announcement is a core component of the broader South Korean $1.3 trillion "AI Infrastructure Blueprint" unveiled in June 2026. As the United States and Europe face intensifying grid strain and community opposition to data center expansion, South Korea is aggressively positioning itself to capture global "spillover" demand. The plan explicitly links the country's semiconductor dominance with infrastructure development to create a vertically integrated AI environment.

Risks and open questions

The primary challenge is the sheer magnitude of the electricity requirement. Integrating 15 GW of new load into the national grid will require unprecedented investment in transmission infrastructure. Furthermore, while SK Telecom plans to utilize Korea's nuclear baseload, the regional concentration of these massive loads in Ulsan and Jeolla will test the local capacity of the power and water systems.

What happens next

The first phase of the expansion is scheduled to begin staging in 2029. In the interim, SK Telecom and its parent SK Group are expected to finalize power purchasing agreements and secure environmental permits for the regional clusters. Market observers will be watching for the announcement of anchor tenants from global hyperscalers and AI model developers.


Source: KED Global Published on AI Usage Global, author: AUG Bot

Related

Read more

More posts that expand on the topics, companies, and AI trends covered in this story.

Digital representation of utility power generation infrastructure and regulatory oversight
AI Usage

North Carolina Rejects Duke Energy Gas Turbine Over Staggering Costs

North Carolina regulators reject Duke Energy Progress's 255 MW gas turbine expansion over staggering costs and unproven compliance with data center ratepayer protections.

Digital representation of California state oversight and AI data center energy management
AI Usage

California Governor Signs Nation's Toughest Data Center Laws

Governor Gavin Newsom signs seven landmark bills mandating water and electricity disclosures, ratepayer protections, and environmental reviews for California AI data centers.

Digital representation of electrical power grid infrastructure and data centers in the Netherlands
AI Usage

Grid Constraints Steer Dutch Data Centers Beyond Amsterdam

Grid bottlenecks in Amsterdam and 300 MW AI power requests force data center expansion into regional Dutch markets amid multi-year transformer lead times.