SK Telecom at MWC 2026: Inside the Carrier's Bet on an AI-Native Future

Korea's largest carrier is rewriting its billing systems, building a gigawatt-scale data centre network, and upgrading its AI model to one trillion parameters — all at once.

BARCELONA — At MWC 2026, most carriers came to showcase new handsets and network specifications. SK Telecom came to announce a different kind of company. Korea's dominant mobile operator used the Barcelona stage to outline what it is calling an "AI Native" strategy — a wholesale restructuring of how its infrastructure, customer systems, and internal culture are built and operated.

The announcement was made by CEO Jung Jai-hun at a press conference during the event, and it covered everything from network automation and a sovereign large language model to hyperscale data centres and a partnership with OpenAI. Taken together, the strategy reads less like a product roadmap and more like a national infrastructure plan.

What "AI Native" Actually Means for a Telecom Operator

The phrase "AI Native" has become a familiar piece of conference vocabulary. At SK Telecom, it appears to carry a specific technical meaning: the wholesale replacement of legacy IT systems with architectures designed from the ground up to be managed and optimised by AI.

According to the company's briefing materials shared at MWC 2026, this includes redesigning sales, line management, and billing systems so they can generate personalised plans and memberships based on each individual customer's usage behaviour — a shift from static pricing tiers to dynamic, data-driven offers.

The company also announced plans to apply a Zero Trust security framework across its systems, incorporating stronger authentication, network segmentation, access controls, and AI-based threat monitoring. For a carrier that handles data for tens of millions of subscribers, that is a significant compliance and architectural commitment.

CEO Jung framed the shift in terms that go beyond operational efficiency.

"SKT is currently at a golden time of transformation, where the two tasks of 'customer value innovation' and 'AI innovation' intersect in a borderless, converged environment that goes beyond telecommunications," Jung said. "SKT defines 'the customer as the very essence of our business,' and through innovation driven by AI, we will evolve into a company that makes meaningful contributions to our customers and to Korea."

Autonomous Networks: AI-RAN and Self-Managing Infrastructure

Alongside the IT overhaul, SK Telecom is expanding what it calls an "autonomous network operations" strategy. The company said it will use AI to automate wireless quality management, traffic routing, and physical equipment operations — functions that have traditionally required large teams of network engineers.

Central to this effort is AI-RAN (AI-enabled Radio Access Network) technology. AI-RAN integrates machine learning directly into base station operations, allowing the network to optimise signal quality, manage interference, and allocate spectrum in real time without human intervention. SK Telecom says the technology aims to improve both download speeds and latency figures — two metrics that directly affect enterprise customers deploying wireless-dependent applications.

For enterprise decision-makers watching the telecom sector, the broader implication is significant. Carriers that succeed in automating network operations at this level would be able to respond to fault conditions and capacity demands faster than any manually managed system could, while also reducing operating costs substantially.

A Single AI Agent Across All Customer Touchpoints

SK Telecom is also redesigning how customers interact with its services. The company plans to simplify pricing, roaming, and membership features and deploy what it calls an integrated AI agent to connect experiences across its main customer portal, T world, and its online store, T Direct Shop.

According to the company, the agent is designed to analyse daily usage patterns and offer tailored recommendations across multiple channels — web, app, and phone. The company is also expanding its AI Contact Center so that human support representatives can use AI-generated prompts and reference material during live customer calls, rather than relying on static scripts.

Offline retail is included in the plan. SK Telecom said AI tools will help store staff identify likely customer needs and follow up with personalised recommendations after a visit. The company is also building what it calls "AI Personas" — customer segment models that analyse digital behaviour patterns and support conversational Q&A for both staff and automated channels.

This mirrors a wider pattern in enterprise service design. The gap between reactive call centre operations and predictive, personalised service is where carriers are now competing. Embedding AI as the connective layer across billing, support, and retail is how SK Telecom is choosing to close it.

Building a 1GW-Class AI Data Centre Network

The most capital-intensive element of SK Telecom's strategy is its plan to build hyperscale AI data centres across Korea, targeting aggregate capacity that exceeds one gigawatt. The company says it aims to attract global investment and position the country as a major AI data centre hub in Asia — a market where Singapore, Japan, and Malaysia are all competing for the same capital flows.

SK Telecom already operates a GPU cluster called Haein and has applied its virtualisation platform, Petasus AI Cloud, to provide GPU-as-a-service workloads for enterprise customers. The company now plans to make Petasus available internationally.

One data centre in Korea's southwestern region will be developed in collaboration with OpenAI, according to the announcement at MWC 2026. The partnership signals a deeper alignment between the two companies than a standard API licensing agreement — though the specific terms of the infrastructure collaboration were not disclosed at the press conference.

Jung used pointed language to describe the infrastructure's national significance:

"AIDC can be seen as the heart of Korea, and hyperscale LLMs as the brain. By combining SKT's AI capabilities with collaboration from domestic and global partners, we will lead true AI-native transformation for Korean customers and enterprises."

From 519 Billion to One Trillion Parameters: Upgrading Korea's Sovereign LLM

SK Telecom's sovereign AI foundation model currently stands at 519 billion parameters, which the company claims is the largest in Korea. The upgrade roadmap targets more than one trillion parameters, with multimodal capabilities — covering image, voice, and video input — due to be added from the second half of 2026.

For enterprise users, the relevant question is not parameter count alone — it is how the model will be applied in sector-specific contexts. SK Telecom gave one concrete example: a manufacturing AI package being developed with SK hynix, a global semiconductor manufacturer, that processes production line data in real time to detect defects and improve equipment efficiency. The package will be offered as a bundled infrastructure, model, and application stack — a vertical AI product rather than a general-purpose API.

Internal Culture: 2,000 AI Agents Already Deployed Internally

Perhaps the most telling indicator of how seriously SK Telecom is taking this transformation is what is happening inside the company itself. The operator has built an AX Dashboard — an internal platform that tracks AI tool usage across departments and at the individual employee level.

It has established an AI Board to govern AI transformation efforts, and created an internal environment it calls the "AI Playground," where employees can build their own AI agents without writing code. According to figures shared at the press event, more than 2,000 AI agents are already in active use across functions including marketing, legal, and public relations.

"To drive future growth, we must reinvent our way of working from the ground up. SKT will fundamentally transform its corporate culture to be centred around AI," Jung said.

For other enterprises, the structural lesson here is less about branding and more about governance. SK Telecom is creating visible accountability for AI adoption — tracking it, governing it through a dedicated board, and giving employees the tools to build agents themselves. That combination of measurement and enablement is harder to replicate than any single technology deployment.

What This Signals for the Broader Industry

SK Telecom's MWC 2026 announcements represent a coherent, if ambitious, attempt to make AI the operating model of a major carrier — not a feature layered on top of existing infrastructure. The combination of rebuilt billing systems, autonomous network operations, a sovereign frontier model, and a hyperscale data centre build-out is a significant bet on vertical integration in an industry that has historically sourced its technology stack from a small number of global vendors.

Whether SK Telecom can execute at the scale it describes remains an open question. Building beyond one gigawatt of data centre capacity while simultaneously retraining a workforce of tens of thousands and upgrading a core LLM to one trillion parameters are each multi-year programs. Doing all of them at once, in a competitive semiconductor supply environment, is a formidable operational challenge.

What is clear from Barcelona is that AI is no longer being positioned as a side investment at SK Telecom. It is the company's stated reason for existing in its next form.

Sources: SK Telecom press briefing at MWC 2026, Barcelona, Spain. March 2026.