Samsung wins two high‑profile AI‑RAN contracts in South Korea
In a joint announcement on 23 September 2026, Samsung Electronics said it has been chosen by both KT Corp and SK Telecom to deliver the network backbone for their AI‑RAN (Artificial‑Intelligence Radio Access Network) trials under the Korean Ministry of Science and ICT’s Hyper AI Network programme. Samsung will act as the sole global supplier for KT’s deployment and as the lead vendor for SK Telecom’s pilot, making it the only company to be selected for both projects.
The contracts are part of a broader government effort to accelerate AI‑driven connectivity, with the ultimate aim of smoothing the transition from today’s 5G to a future 6G ecosystem. For UK consumers and contract‑deal hunters, the initiative signals that the next generation of ultra‑reliable, low‑latency services – from autonomous vehicles to remote‑controlled robotics – could become mainstream much sooner than previously thought.
What the pilots will look like
KT’s shipyard test‑bed
KT will roll out a private 5G Standalone (SA) network at the Yeongam shipyard of HD Hyundai Samho. The site will host a suite of physical‑AI services, the most eye‑catching being a quadruped welding robot that can move around the hull and perform precision welds without human intervention.
Samsung’s contribution centres on its Network‑in‑a‑Server (NIS) solution – an edge‑centric platform that bundles a virtualised RAN (vRAN), an AI‑native core and the necessary AI applications. NIS runs on a commercial‑off‑the‑shelf Supermicro Compact Edge Server powered by an AMD EPYC 8004 CPU and the Wind River Cloud Platform. The architecture is deliberately modular, allowing GPU accelerators to be added when a particular use‑case demands extra compute power.
The edge server sits on‑premises, meaning data never has to travel to a distant data centre. This proximity is crucial for the welding robot, which requires sub‑10 ms latency, high reliability and steady throughput to keep the control loop tight and avoid any safety‑critical delays.
“KT views this project as an important starting point for validating the next‑generation communications paradigm where AI and networks converge,” said Jongsik Lee, EVP at KT.
SK Telecom’s petrochemical safety showcase
SK Telecom’s trial will be hosted at SK Incheon Petrochem, a large industrial complex that supplies much of the Seoul metropolitan area. Here the focus shifts to autonomous patrol robots that navigate hazardous zones, streaming high‑definition video back to a control centre for real‑time AI analysis.
Again, Samsung will supply its NIS platform, this time to prove that AI‑native connectivity can improve both safety and operational efficiency in environments where a single mistake can have severe consequences. The robot’s live video feed will be processed on‑edge, allowing AI models to flag anomalies such as leaks or equipment malfunctions instantly.
Takki Yu, head of SK Telecom’s Network Technology Office, summed up the ambition: “AI creates greater value when paired with the right foundation – this project puts that belief into practice.”
Why the UK should pay attention
While the pilots are firmly rooted in Korean industry, the technology stack they showcase is global. Samsung’s NIS platform is built on open standards and can be deployed on any compatible COTS hardware, meaning the same solution could be rolled out in UK factories, ports or even large‑scale events.
For Android‑phone buyers hunting for the best contract deals, the relevance lies in the network quality that will eventually trickle down to consumer services. AI‑enhanced RAN promises:
- Dynamic traffic optimisation – AI can predict congestion and re‑allocate spectrum in real time, keeping download speeds stable even in crowded city centres.
- Energy efficiency – smarter radio scheduling reduces power consumption, potentially lowering operator‑wide OPEX and, in turn, consumer tariffs.
- Improved coverage – AI‑driven beamforming can extend reliable 5G reach into traditionally hard‑to‑cover areas such as underground stations or dense high‑rise districts.
If the Korean trials prove successful, UK operators are likely to adopt similar AI‑RAN tools to stay competitive, especially as the 6G roadmap begins to take shape. Early adopters could market “AI‑optimised” plans, a new selling point for tech‑savvy customers.
Technical deep‑dive: NIS and CognitiV NOS
Samsung’s CognitiV Network Operations Suite (NOS) works hand‑in‑hand with NIS, providing a single pane of glass for network automation, anomaly detection and performance analytics. The suite leverages machine‑learning models trained on live traffic to automatically adjust parameters such as MIMO layer count, carrier aggregation and power levels.
Key capabilities highlighted in the press release include:
- Anomaly detection – spotting unusual latency spikes or packet loss before they impact services.
- Throughput enhancement – AI‑guided scheduling that squeezes extra megabits out of existing spectrum.
- Energy savings – predictive sleep‑mode activation for base‑band units during low‑traffic periods.
All of these functions run on the same edge hardware that hosts the vRAN, meaning there is no need for separate analytics clusters – a cost‑effective model that could appeal to UK enterprises looking to build private 5G networks for factories, campuses or logistics hubs.
Timeline and next steps
- October 2026 – Both pilots go live, with Samsung delivering the NIS hardware and software stack.
- 2027 – Field trials will generate performance data, feeding back into Samsung’s AI‑RAN algorithms and informing the Korean government’s AI Highway agenda.
- Beyond 2027 – Successful outcomes are expected to accelerate commercial roll‑outs, both in Korea and internationally, as operators seek to differentiate their 5G offerings and lay the groundwork for 6G.
Bottom line for UK contract shoppers
Samsung’s involvement in these high‑profile AI‑RAN projects underscores a shift from static 5G networks to intelligent, self‑optimising systems. While the immediate impact will be felt in Korean shipyards and petrochemical plants, the technology transfer to global markets is inevitable.
For Android‑phone users, this could translate into more stable 5G connections, lower latency for gaming and AR, and eventually new service tiers that leverage AI‑optimised back‑haul. Keep an eye on announcements from UK carriers – they may soon start touting AI‑enhanced network features as part of their next contract bundles.
Samsung’s AI‑RAN pilots are a clear indicator that the next wave of mobile connectivity will be software‑centric, AI‑driven and ready for 6G. As the ecosystem matures, UK consumers can expect the benefits of these innovations to filter down through their everyday mobile plans.
