AI Inference Moves Into Live 5G Networks

At Mobile World Congress (MWC) Barcelona, taking place March 2–5 2026, Intel presented advancements that demonstrate live AI inference in operational mobile networks and a path toward more efficient and automated 5G and beyond deployments. These developments are relevant to telecommunications operators, cloud providers, and enterprises modernizing wireless infrastructure with AI-optimised network functions.

AI Inference Integrated into Live Networks
Intel is showcasing how AI inference workloads can run within live mobile networks on a unified open platform. By integrating AI functions closer to network edge elements such as the radio access network (RAN) and core network nodes, operators can use real-time analytics to optimise traffic flows, reduce congestion, and enhance signal performance. This aligns with broader industry efforts to embed machine intelligence into 5G Advanced network operations, where AI and machine learning are standardised to improve spectral efficiency and automation.

Through demonstrations at MWC, the company highlights ecosystem cooperation across core network, RAN, enterprise networks, and enterprise edge domains. The emphasis is on scaling next-generation infrastructure without requiring wholesale hardware replacements, aiming to reduce total cost of ownership while improving performance.

Supporting Standards and Industry Collaboration
Intel’s approach builds on decades of work with open standards and partnerships with operators, cloud service providers, and technology vendors. By using common platforms and co-engineering with ecosystem partners, the company asserts that deployments can be standardised and scaled across varied environments, from public mobile networks to private enterprise settings.

This collaborative framework is consistent with industry momentum around programmable and AI-augmented networks, which are expected to support a wide range of applications from enhanced mobile broadband to industrial automation and smart city connectivity where digital supply chain integrity and low-latency communication are critical.

Practical Network Optimisation and Use Cases
Running AI inference in live networks has measurable impacts on operational metrics. When AI processes are located closer to the edge for example within RAN controllers or multi-access edge computing nodes they can respond to dynamic conditions such as fluctuating traffic loads or interference patterns in real time. For end users, this can result in clearer voice and video communication, more consistent throughput, and reduced latency. For network operators, this means greater automation, more efficient use of spectrum and hardware resources, and improved return on infrastructure investment.

At the MWC event itself, Intel is hosting panels and technical discussions with industry experts and showcasing solutions on the exhibition floor to illustrate these concepts in action.

Positioning for Future Generations
While 5G Advanced deployments are still underway globally, integrating AI inference directly into network layers lays groundwork for future network generations, including 6G. By enabling richer automation and intelligence at scale, the industry can address the increasing complexity and data demands of emerging applications across consumer, enterprise and industrial sectors.

www.intel.com