KT, SK Telecom tap Samsung for AI

- Samsung landed key roles in South Korea’s government-backed AI-RAN push
- The projects will deploy private 5G SA networks in industrial sites to test AI-RAN and physical AI use cases
- Samsung said its role with KT and SK Telecom build on decades-long partnerships
It’s not 6G, but Samsung Electronics is well on its way to providing KT and SK Telecom with AI-RAN solutions to advance physical AI on the path to AI-native 6G.
Samsung Electronics this week announced that it signed contracts with KT and SK Telecom for their AI-RAN projects under the “Hyper AI Network” initiative at the Korean Ministry of Science and ICT (MSIT). Samsung will be the sole global vendor for KT and the main vendor for SK Telecom.
It’s a big job for Samsung and one that builds on its decades-long partnerships with the country’s leading operators. Samsung noted that it’s the only partner selected for both projects.
The work is set to begin next month and will deploy 5G Standalone (SA) private networks across industrial environments – ranging from shipyards to petrochemical facilities – to validate AI-RAN capabilities and generate real-world insights for future deployments.
Samsung – plus Ericsson, Nokia, HFR
Recon Analytics analyst Daryl Schoolar said he’s not surprised that both mobile network operators are working with Samsung on this project.
“AI-RAN is one of the key 6G capabilities being promoted by the industry. There is hope that AI-RAN will help make 5G use cases scale easier,” he told Fierce.
However, SK Telecom isn’t just working with Samsung, he pointed out. SK Telecom announced in July that it’s working with four manufacturers – including Samsung, HFR, Ericsson and Nokia – to develop and deploy their AI-RAN equipment for the Hyper AI Network.
Robots do the welding
What makes this AI-RAN announcement stand out from others is the trial includes physical AI in the form of robots for welding, Schoolar said.
For the quadruped welding robot use case, Samsung said it will support KT with its software-driven, AI-native solutions. That includes Network in a Server (NIS) – an integrated edge AI platform comprised of virtualized RAN (vRAN), AI core and AI applications – and Samsung CognitiV Network Operations Suite (NOS) for intelligent network automation.
According to Samsung, the quadruped robot can autonomously perceive its working environment and perform precise welding by leveraging 5G connectivity and the edge AI platform.
For those keeping track, Samsung’s NIS runs vRAN and 5G SA core software on an AMD EPYC 8004 Server CPU-powered commercial off-the-shelf (COTS) Compact Edge Server from Supermicro with Wind River Cloud Platform.
Another part of Korea’s Hyper AI Network initiative is SK Telecom’s role in validating physical AI services at SK Incheon Petrochem, a major energy hub for the Seoul metropolitan area.
In that implementation, an autonomous robot will patrol hazardous areas and transfer high-definition videos in real time while on the move, enabling AI analysis to detect potential hazards and facilitate integrated monitoring.
Either way you look at it, it looks as though robots will be doing the heavy lifting while AI-RAN leads the way.
Read more about Samsung and AI-RAN:
Samsung charts its own path to AI-RAN
MWC: Softbank, Nvidia, T-Mobile and more kick off AI-RAN Alliance

Facts Only

* Samsung signed contracts with KT and SK Telecom for AI-RAN projects under the "Hyper AI Network" initiative at the Korean Ministry of Science and ICT (MSIT).
* Samsung will be the sole global vendor for KT and the main vendor for SK Telecom.
* The projects involve deploying 5G Standalone (SA) private networks in industrial sites to test AI-RAN and physical AI use cases.
* Deployment will occur across industrial environments, including shipyards and petrochemical facilities.
* Samsung's solution includes Network in a Server (NIS), an integrated edge AI platform comprising vRAN, an AI core, and AI applications.
* Samsung also provides the Samsung CognitiV Network Operations Suite for intelligent network automation.
* A quadruped welding robot use case will leverage Samsung’s software-driven, AI-native solutions.
* SK Telecom is validating physical AI services at SK Incheon Petrochem using autonomous robots to patrol hazardous areas and transfer video.
* Samsung's NIS runs vRAN and 5G SA core software on an AMD EPYC 8004 Server CPU from Supermicro with Wind River Cloud Platform.
* Other partners include Ericsson, Nokia, and HFR in the AI-RAN development for SK Telecom.

Executive Summary

Samsung Electronics has secured contracts with KT and SK Telecom for AI-RAN projects under the "Hyper AI Network" initiative at the Korean Ministry of Science and ICT. Samsung will serve as the sole global vendor for KT and the main vendor for SK Telecom on these projects, building upon existing partnerships with the operators. The work involves deploying private 5G Standalone (SA) networks in industrial environments to test AI-RAN capabilities and generate real-world data. This effort is positioned as a step toward enabling physical AI using the AI-RAN framework on the path to AI-native 6G.
The scope of the work includes deploying these private networks across industrial sites, such as shipyards and petrochemical facilities. Samsung will provide software-driven, AI-native solutions, including the Network in a Server (NIS) platform for edge AI and the Samsung CognitiV Network Operations Suite for automation. Furthermore, SK Telecom is involved in validating physical AI services through implementations like an autonomous robot patrolling hazardous areas at SK Incheon Petrochem. Other manufacturers, including Ericsson, Nokia, and HFR, are also collaborating on the development of AI-RAN equipment for the Hyper AI Network alongside Samsung.

Full Take

The narrative positions AI-RAN not merely as a technological upgrade but as a critical foundational layer necessary for realizing physical AI capabilities within future 6G systems. The focus on industrial deployment—using robots for welding and autonomous monitoring in hazardous environments—suggests an immediate, tangible application where the benefits of edge AI processing and real-time network intelligence translate directly into operational safety and efficiency, moving beyond theoretical benchmarks. The structure implies a progression: building foundational AI infrastructure (AI-RAN) allows physical systems (robots) to interact intelligently with that infrastructure to perform complex tasks.
A key area for inquiry is the implied hierarchy of vendor roles. Samsung occupies a unique position as the sole global vendor for KT and the primary vendor for SK Telecom, suggesting a dependency rooted in long-term partnerships, which could influence future industry standards or proprietary integration paths. The inclusion of multiple competitors (Ericsson, Nokia, HFR) working on the broader AI-RAN equipment suggests that while Samsung leads specific aspects, the ecosystem is intentionally multi-vendor. This dynamic warrants examining whether the "Hyper AI Network" initiative is truly an open standard push or a framework where hardware and software consortia converge to achieve a shared outcome.
The pattern of linking cutting-edge network technology (AI-RAN) with physical automation (robotics) reflects a systemic shift toward cyber-physical systems where connectivity, intelligence, and physical action are inseparable. The implication is that the value is not solely in the bandwidth or the network slicing, but in the closed-loop intelligence that allows machines to perceive, decide, and act within complex industrial spaces. This focus risks obscuring the potential costs associated with deploying such integrated systems, specifically concerning data security, regulatory oversight of autonomous physical actions, and the equitable distribution of AI-derived operational advantages across the industrial sector.
Bridge Questions: What metrics are being used to measure the efficacy of the AI-RAN validation in terms of operational savings versus deployment complexity? How is responsibility allocated when AI-driven robotic actions occur in environments validated via these networks? What are the long-term implications for regulatory frameworks concerning autonomous physical systems operating on shared telecommunication infrastructure?

Sentinel — Human

Confidence

The article reads as a well-researched news report that effectively synthesizes contractual information with technical context and expert commentary.

Signals Detected

Sentence length variance shows natural variation; sophisticated topic handled with direct quotes.

Smooth transition between specific contracts, analyst commentary, and technical details without overt hedging or unnatural balancing.

Attribution of the primary contract to Samsung/KT/SKT is direct; secondary statements are sourced (Daryl Schoolar) or attributed to the company, avoiding verbatim template repetition.

Specific technical details (NIS components, COTS servers) are present and contextually consistent, suggesting grounding in specific industry knowledge rather than pure LLM fabrication.

Human Indicators

The inclusion of specific, deep-dive technical architecture (vRAN, NIS, AMD EPYC CPUs) layered alongside high-level partnership announcements suggests input from an insider or deeply researched journalist.

The flow successfully shifts between corporate announcement, analyst perspective, and physical application examples (welding robots), demonstrating a narrative weaving characteristic of human synthesis.