Telecom companies, leveraging extensive real estate and mature network infrastructure, are rethinking their role as the world transitions into an AI-first era.
Enterprise leaders face a growing challenge: the compute and energy requirements of generative AI models are rapidly outpacing traditional data centre capacity. As organisations move beyond pilot projects, they encounter rising costs, latency problems from centralised cloud setups, and mounting pressure to meet sustainability targets.
This infrastructure gap has opened opportunities for new models and entrants. Telcos are positioning themselves to move beyond pure connectivity providers toward becoming foundational components of the AI value chain.
SK Telecom (SKT) has launched a concerted effort to capitalise on that shift. The South Korean operator announced it is “evolving into a comprehensive artificial intelligence data centre (AIDC) developer” and is preparing to expand into international markets.
At the SK AI Summit 2025, newly appointed CEO Jung Jaihun outlined a strategy that casts SKT as not merely a consumer of AI services but as a builder of the underlying infrastructure that supports them.
SK Telecom sets out plan for global enterprise AI infrastructure
SK Telecom’s strategy rests on two main pillars: creating a robust domestic AI infrastructure backbone and exporting its AIDC model overseas.
Domestically, SKT is developing AIDC hubs across Korea’s three key regions: the Seoul metropolitan area, the southern region around Ulsan, and the southwest region. The company intends to attract global investment and position South Korea as a major Asian AI hub. The Ulsan AIDC is central to the plan, with ambitions to scale it to a 1 GW-class facility through international partnerships.
For organisations outside Korea, SKT’s international expansion is particularly relevant. The company plans to enter Southeast Asia with “energy-specialised AIDC solutions” built together with SK Group affiliates.
One concrete example is a planned AIDC in Vietnam, developed in collaboration with SK Innovation. The facility is designed to secure a stable power supply from a liquefied natural gas (LNG) power plant and to reuse the plant’s regasification “cold energy” for data centre cooling. By addressing power and cooling—the two largest operational costs for AI-heavy infrastructure—SKT aims to create a cost- and energy-efficient model.
SKT is also acting as its own anchor customer. The company plans to acquire more than 2,000 NVIDIA RTX PRO 6000 Blackwell GPUs to build a Manufacturing AI Cloud. CEO Jung said this cloud will support AI advancements across SK Group’s manufacturing businesses, such as SK Hynix, powering use cases like digital twins and robotics AI.
Implementation, challenges, and ecosystem
SKT’s approach leverages the company’s telco identity to build data centres differently. CEO Jung argues that telecom networks are being reassessed for their new strategic value in an AI-driven world.
“Using nationwide telecom infrastructure, telecom operators can uniquely bridge central AIDCs and on-device AI through edge AI and intelligent telecom technologies such as AI-enabled RAN,” Jung said.
The proposed architecture is hybrid: large centralised AIDCs would handle training of large models, while telco edge networks would deliver low-latency inference for on-device applications. Realising this vision relies on a broad partner ecosystem and close alignment with major cloud and hardware vendors.
On the cloud side, SKT has formed an R&D partnership with Amazon Web Services (AWS) to accelerate Edge AI deployments, building on a prior memorandum of understanding for the Ulsan AIDC. On the hardware front, SKT is collaborating with NVIDIA along with government and academic partners to trial AI-enabled RAN technology that integrates AI capabilities into the network itself.
To serve enterprise customers, SKT intends to move beyond basic colocation and become what Jung describes as a “comprehensive AIDC developer,” handling design, construction, and operation of AIDC projects end to end.
SKT plans to commercialise an AIDC solution package that bundles internalised technologies for cooling and power management (including out-rack and energy solutions), server interconnection through clustering solutions, and server efficiency measures (in-rack optimisations).
The future of AI infrastructure sourcing
For CIOs, CTOs, and COOs, SK Telecom’s moves offer several lessons about how AI infrastructure will be sourced going forward.
First, specialised providers are on the rise. Hyperscalers such as AWS, Google, and Microsoft will remain dominant, but telcos—with strengths in distributed edge locations, network control, and energy management—represent a complementary partner class, especially for energy-intensive or low-latency workloads.
Second, infrastructure is becoming a competitive advantage. As CEO Jung noted, “AI infrastructure is a core driver of both corporate and national competitiveness.” Access to efficient, scalable, and sustainable compute is now a strategic capability that influences product innovation, time to market, and operational cost — elevating infrastructure decisions to the boardroom level.
Finally, SKT’s ambition to build an AI hub in Korea underscores the growing emphasis on data sovereignty and regional hub strategies. For organisations navigating regulatory constraints, regional AIDCs that support sovereign data policies for training and inference will become increasingly attractive. Leaders should align AI and data governance policies with these emerging infrastructure options.
Overall, SK Telecom’s initiative signals that the industry is actively seeking new answers to the economic and engineering challenges posed by enterprise-scale AI deployment.
See also: How AI-native 6G networks will boost enterprise operations
Want to learn more about AI and big data from industry leaders? Check out the AI & Big Data Expo, held in Amsterdam, California, and London. The event is part of TechEx and is co-located with other technology conferences including Cyber Security Expo.
Telecoms is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars through TechForge Media’s event listings.