SK Telecom Outlines 1GW AI Data Center and 1T LLM Strategy

Barcelona, Spain – SK Telecom announced plans to build a nationwide one-gigawatt infrastructure for artificial intelligence data centers and to upgrade its proprietary large language model to more than one trillion parameters, signaling a strategic shift toward AI-integrated operations.

Chief Executive Jung Jai-hun outlined the strategy at a press conference in Barcelona.

The infrastructure plan includes establishing a data center in South Korea's southwestern region in partnership with OpenAI, with the objective of positioning the country as a regional hub for AI computing.

SK Telecom stated it will collaborate with affiliates SK Hynix, SK Ecoplant, and SK Innovation to manage the supply chain, including construction, cooling systems, servers, and energy operations.

The company plans to expand its Korean-language sovereign AI model from 519 billion to more than one trillion parameters and to add multimodal capabilities for voice and video processing in the second half of the year.

SK Telecom also plans to market its virtualization platform, Petasus AI Cloud, internationally, following its initial deployment on the Haein GPU cluster.

In its telecommunications business, SK Telecom plans to migrate integrated IT systems—including sales, billing, and line management—to AI-driven architectures secured by a Zero Trust framework.

The company also plans to introduce AI-based autonomous network operations for wireless quality, traffic management, and equipment oversight using AI-RAN technology.

For consumers, SK Telecom is developing a unified AI agent across its digital platforms, including T World and T Direct Shop, and is enhancing its A. phone application to automate scheduling and call management.

In the enterprise segment, SK Telecom and SK Hynix are developing an AI solution package intended to reduce defect rates and improve equipment efficiency in the manufacturing and semiconductor sectors.

Internally, SK Telecom has deployed more than 2,000 AI agents created by employees across various departments using a no-code development platform, supported by a dashboard designed to monitor AI adoption across the organization.