[Edaily Reporter Kim Hyun-ah ] SKTelecom(017670)The company is shifting its focus from a telecommunications-centric business structure to AI, achieving successive successes in the government’s major AI projects.

The company is expanding its business scope beyond the development of its own AI models to include AI services for the general public, industrial and security AI, and AI data centers (AIDC).

The company has also restructured its organization. It has introduced a “dual MNO and AI CIC” system, operating telecommunications and AI as separate in-house companies (CICs), to enable different decision-making processes and investment paces based on the characteristics of each business. Under this structure, the core telecommunications business focuses on a stable customer base and service competitiveness, while the AI division enhances agility to adapt to rapidly changing technologies and markets.

Ranked

First in the Second Phase of the “Dokpamo” Evaluation… Making Its Presence Felt in the “Model Sovereignty” Competition

The most notable achievement is the government’s “Independent AI Foundation Model (Dokpamo)” project.

In the second-stage evaluation released by the Ministry of Science and ICT on the 27th, SKT’s elite team secured first place among participating teams with a total score of 70.6 points. It outperformed Upstage (69.9 points) and LG Corp. (69.0 points). This evaluation was conducted by synthesizing domestic benchmarks, expert evaluations, professional user evaluations, and general user evaluations.

SKT demonstrated strengths in domestic benchmarks—which assess Korea’s linguistic, cultural, and historical contexts—as well as in user evaluations designed to verify practical applicability. It is assessed that the company demonstrated the capability to develop AI models that can be practically utilized in the domestic user environment, going beyond a simple performance competition with global models.

SKT has been enhancing its AI model competitiveness centered on its in-house developed ultra-large AI model series, “A.X K.” By accumulating experience in developing large-scale models—such as “A.X K1,” which features 500B (500 billion) parameters—the company is also participating in the “Sovereign AI” competition aimed at reducing dependence on global Big Tech.

Beyond Models to “AI for the Public”… Selected as a “AI for All” Service Provider

The competitive edge secured in AI

models

is now expanding into the service domain.

On the 28th, SKT was officially selected as an implementing agency for the “Support for Universal Adoption of AI Services for All Citizens (AI for All)” project promoted by the Ministry of Science and ICT. Among six competing consortia, three—including SKT, Kakao, and KTCorporation—were selected as project operators.

The vision proposed by SKT goes beyond chatbots that simply answer questions to “AI that does the work for you.”

The vision is to implement AI agents that find information needed in daily life—such as job hunting, financial and household management, and accessing public services—and connect users to follow-up tasks.

In addition to its proprietary AI models, SKT can leverage the Agent AI technology accumulated through “A.dot” and its mobile subscriber base of 22.5 million as a foundation for expanding AI services.

The goal is not to stop at simply creating AI models, but to build a business structure that progresses from models to Agent AI to services.

The Next Stage: Security AI… Combining ‘Models and Security’

The next test will be the “Security-Specialized AI Foundation Model (Specialized AI Foundation Model)” business.

A competitive landscape has emerged, with the SKT-Upstage consortium and the Naver Cloud-LG Corp. camp both throwing their hats into the ring.

The SKT consortium includes SKT and Upstage, both of which passed the second round of evaluation for the Specialized Security Foundation Model (SSFM) project. They have been joined by major domestic security firms such as SK Shields, AHNLAB,INC.(053800), SecuI, IGLOO Corporation, and RaonSecure Co., Ltd.(042510).

The plan is to combine foundation model development capabilities with data and service capabilities from the cybersecurity field to implement an “agent-based security” system in which AI detects and responds to cyber threats.

Since the results of the operator selection process have not yet been announced, it is difficult to predict the outcome. However, the initiative is drawing significant attention as it brings together AI model developers that have proven their competitiveness in the “Dokpamo” competition and major domestic security firms.

Two Major CICs: Telecommunications and AI… AI Data Centers Also Emerge as a Growth Engine

SKT’s transformation extends beyond technology and services. The company is restructuring its organization to adapt to the AI era.

In its November 2025 organizational restructuring, SKT reorganized its structure around two core business units (CICs): Telecommunications (MNO) and AI. The MNO CIC focuses on restoring competitiveness in its core telecommunications business and securing customer trust, while the AI CIC focuses on driving business initiatives with flexibility and speed to generate tangible results.

The company plans to adopt different approaches to decision-making and business operations, taking into account the distinct characteristics of the telecommunications business—where stable service quality and customer base management are crucial—and the AI business, which is characterized by rapid technological change.

The AI CIC is strengthening not only its B2C and B2B AI businesses but also its data center operations. Last April, the AIDC business unit within the AI CIC was expanded and reorganized into the “AI DC Business Division” and the “AIDC Development Division.” This move is intended to cultivate AI infrastructure as a new growth engine.

In particular, the AI DC business is being reorganized to establish a three-pronged approach—strategy, operations, and new development—with roles divided among SKT, SK Horizon, and SK Hyper. SKT will be responsible for overall business strategy and global collaboration, SK Horizon will handle the operation and expansion of existing data centers, and SK Hyper will focus on the development of large-scale new AI DCs.

SKT has secured management control by retaining a 51% stake in SK Horizon while attracting a total of 3.08 trillion won in external investment. Based on this, MIRAE ASSET SECURITIES expects the data center business expansion to accelerate and has raised SKT’s target stock price by 25%, from 120,000 won to 150,000 won.

By differentiating the operational approaches for each business—maintaining the subscriber, network, and data capabilities secured through telecommunications under stable MNO operations, while allowing AI and AIDC to move more quickly in response to technological and market changes—the company has established a differentiated operational structure.

From a ‘Telecom Company’ to an AI Technology Firm

The keywords defining SKT’s recent trajectory are ‘expansion’ and ‘connection.’

If Dokpamo serves as the stage where the competitiveness of AI models is validated, “AI for All” represents the stage of expanding these models into services used by the general public. Dokpamo extends AI into the realms of national cybersecurity and industry, while AIDC is responsible for the infrastructure that supports these efforts.

This strategy involves connecting AI models, Agent AI, services, and data centers into a single AI value chain, rather than treating them as separate businesses.

Given the rapid pace of technological change in the AI market, not only technical capabilities but also the speed of decision-making and investment are critical. Separating the MNO and AI into a CIC and expanding new growth areas—such as AIDC—within the AI CIC is a strategic choice designed to allocate resources according to the characteristics of each business and enable rapid responses.

Of course, it is difficult to conclude that SKT’s AI competitiveness has been fully proven based solely on the results of government projects. Whether the AI business will generate profitability and growth in the actual market remains a separate challenge.

However, it is clear that the company’s strategic focus is shifting away from the traditional competition centered on telecommunications networks and subscribers toward developing AI models, integrating them with agents and services, expanding into industrial and national sectors, and securing the infrastructure required to power AI.

Whether SKT will establish itself as a technology company that connects AI models, services, and infrastructure beyond telecommunications will ultimately be determined by its performance in the market.