The Vision for an AI‑Centric Network
SK Telecom is moving beyond simple AI add‑ons; it is redesigning its core from the ground up. The company will embed machine‑learning intelligence into every layer of its network, ensuring that routing, optimisation and fault detection are all powered by predictive models.Re‑writing Internal Systems
To support this vision, SK Telecom is rewriting legacy software stacks. By adopting open‑source frameworks and containerised micro‑services, it aims to reduce latency and increase flexibility, allowing the network to learn and adapt in real time.Gigawatt‑Scale Data Centres
The company plans to expand its data‑centre capacity to a gigawatt scale. This expansion will house massive storage clusters and specialised AI accelerators, giving SK Telecom the compute muscle needed to train and host large language models.Upgrading the In‑House LLM
A key element is the upgrade of SK Telecom’s own large language model. The model will be fine‑tuned on telecom‑specific data, enabling it to handle complex customer queries, optimise network traffic, and even generate network code.Revolutionising Customer Experience
AI will also power the next generation of customer‑service desks. From conversational agents that understand nuanced user needs to predictive maintenance alerts, the aim is to provide seamless, personalised support across all channels.Why This Matters for the Industry
SK Telecom’s strategy signals a new era where operators not only consume AI but build it in-house. This shift could accelerate 5G rollouts, improve network reliability, and unlock new revenue streams through AI‑driven services.Take the Survey
Want to share your thoughts on AI in telecom? Take our quick survey here: https://dakik.co.uk/surveyWritten by Erdeniz Korkmaz· Updated Mar 2, 2026
