"Telecom companies must go beyond delivering data quickly and safely to become the designers and drivers of AI infrastructure." (Jung Jai-hun, SK Telecom president)
"We must evolve KT's identity into an AI transformation (AX) platform corporations." (Park Yun-young, KT president)
"We will build a state-of-the-art data center that meets global standards to seize the market." (Hong Bum-Shik, LG Uplus president)
The 2026 shareholders meeting, which offered a glimpse into telecom companies' management direction this year, has wrapped up. With leadership changes completed at SK Telecom and KT, they have started securing future growth engines. Jung Jai-hun, SK Telecom president; Park Yun-young, KT president; and Hong Bum-Shik, LG Uplus president, cited AI-centered B2B (corporations-to-corporations transaction) such as expanding AI data centers (AIDC) and AI contact centers (AICC) as the battleground for this year, aiming to break through the growth limits of the saturated telecom market.
In fact, the area showing growth in telecom companies' results is B2B businesses centered on AI. Last year, SK Telecom's AI data center and AX (AI transformation) business institutional sector revenue was 718.5 billion won, up 25.6% from a year earlier. Over the same period, mobile communications revenue was 1.2051 trillion won, down 5.7% year over year.
In KT's case, revenue in the AI·IT business institutional sector based on AX platforms—such as AICC, Internet of Things (IoT), cloud and smart mobility—rose 3% last year from a year earlier, while KT's B2C (corporations-to-consumer transaction) revenue increased 1.9% over the same period. LG Uplus also saw its AI DC revenue climb to 422 billion won last year, up 18.4% from a year earlier, but B2C revenue rose only 3.3% in the same period.
Jung Jai-hun, SK Telecom president, said at MWC 2026, the world's largest mobile exhibition, on Mar. 2 (local time), "The unique infrastructure and operational know-how of telecom companies are the key to building AI infrastructure and spreading services," adding, "Competitiveness in the AI era depends not only on technology but also on how you design and consolidation the infrastructure."
SK Telecom is currently partnering with Amazon Web Services (AWS) to push ahead with the Ulsan AI data center. Ground was broken in September last year, and it also plans to work with OpenAI to build an ultra-large AI data center in the southwest region. In addition, it successfully advanced to phase two of the government's "national AI model" project and moved to secure sovereign AI competitiveness through its self-developed AI model (A.X K1).
Park Yun-young, KT president, expressed ambition in an inaugural address on Mar. 31 to "evolve KT's identity into an AX platform corporations." Park reflected this direction directly in a reorganization. In the latest reshuffle, KT integrated functions dispersed across B2B strategy, business, technology and partnerships and created a new "AX business institutional sector." The company recruited Park Sang-won, a 1968-born executive director and former consultant, to head the institutional sector. The existing CTO organization, the Technology Innovation Division, was split into the AX Future Technology Institute (CTO) and the IT Division (CIO). Choi Jeong-gyu, head of the Agentic AI Group (managing director) at LG AI Research Institute, was named to lead the AX Future Technology Institute. Under Choi will be three research organizations: Frontier AI Lab, Agentic AI Lab and AX Data Lab. Kim Jun-seok, head of AI at Hanwha Life Insurance who developed Naver's AI-based translation service Papago, was recruited to lead the Agentic AI Lab. The AX Data Lab will also be led by an external hire, Managing Director Lee Sang-bong. The Frontier AI Lab will be led by Managing Director Park Jae-hyung.
Starting with the independently developed large language model (LLM) "Mit:eum 2.0," KT is accelerating its B2B business with a multi-LLM strategy by unveiling models such as the AI "SOTA K," developed in collaboration with Microsoft (MS), and Lama K, developed on an open-source basis. In particular, SOTA K is a Korea-style AI model that KT launched targeting domestic corporations' AX demand, based on the ChatGPT 4o model. In collaboration with MS, the company introduced Secure Public Cloud, a cloud solution specialized for security, in Nov. last year. It is also partnering with the U.S. company "Palantir." KT plans to combine KT Cloud and network infrastructure with Palantir's core solutions to provide services optimized for the Korean market.
Hong Bum-Shik, LG Uplus president, also passed an agenda item at the 2026 shareholders meeting held on Mar. 24 to add "data center design·build·operate (DBO) related operations and services·construction business" to the company's business purposes to expand the B2B AX business. The DBO business is an integrated consulting business that, under commission from client companies, handles the entire process from data center design to construction and operation, laying the groundwork for expanding the data center business. LG Uplus is currently investing 615.6 billion won in Paju, Gyeonggi Province, to build an AI data center with completion targeted for 2027. There, the company plans to provide GPUaaS (GPU subscription service) based on ultra-high-performance graphics processing unit (GPU) servers.
Hong said, "The core competitiveness of the LG Uplus AI data center is the 'infinite LG synergy' that brings together the group's capabilities, such as LG Electronics' next-generation liquid cooling solution and LG Energy Solution's stable power infrastructure technology," adding, "We have also secured expertise and reliability with know-how accumulated from operating data centers for 28 years."
Global market research firm Fortune Business Insights estimated last year's AI data center market at $17.7 billion (26.5924 trillion won) and projected it will expand to $93.6 billion (140.5684 trillion won) by 2032, recording an average annual growth rate of 26.8%.
Hong In-gi, a professor in the Department of Electronic Engineering at Kyunghee University, said, "Telecom companies managed to hold on in B2C by securing 4G (fourth-generation mobile communication) and 5G (fifth-generation mobile communication) subscribers, but they have hit growth limits due to population decline and the entry into an aging society," while adding, "However, as AI boosts corporations' demand to build infrastructure based on 5G, they are trying to generate revenue on the B2B side."