Ericsson Unveils AI-Ready Radios and RAN Software Ahead of MWC 2026

Ericsson has introduced a major new portfolio of AI-ready network infrastructure products ahead of Mobile World Congress 2026 in Barcelona, signaling the telecom industry’s growing focus on supporting the next generation of artificial intelligence and augmented reality applications.

The new lineup includes advanced radios, antennas, and AI-powered Radio Access Network (RAN) software designed to help communications service providers manage rapidly increasing demands from AI-enabled smartphones, connected devices, and immersive digital experiences.

As AI applications become more data-intensive and interactive, telecom networks are facing new technical challenges. Emerging technologies such as multimodal AI assistants, real-time augmented reality, AI-generated media, and edge computing require significantly faster uplink performance, lower latency, and more intelligent network optimization than traditional mobile traffic.

To address these evolving requirements, Ericsson unveiled a portfolio built around three core pillars: AI-ready radios, AI-powered RAN software, and high-performance antennas.

One of the biggest highlights is Ericsson’s new generation of Massive MIMO radios powered by upgraded Ericsson Silicon technology. These systems now include integrated neural network accelerators specifically optimized for artificial intelligence and machine learning workloads.

The neural network accelerators use programmable matrix-core architecture to improve AI inference capabilities directly inside network hardware. This enables telecom networks to perform real-time optimization and dynamic traffic management more efficiently while supporting increasingly complex AI-driven applications.

Ericsson introduced ten new AI-ready radios within the portfolio. These include next-generation Massive MIMO systems engineered to improve both downlink efficiency and uplink performance — a key requirement for AI and AR services that continuously send large amounts of data back to cloud systems.

The company also expanded its high-power radio offerings designed for network consolidation and spectrum efficiency, alongside new TDD Massive MIMO products capable of delivering higher bandwidth and stronger overall network performance.

In addition to the hardware launch, Ericsson introduced several AI-powered RAN software features intended to improve network intelligence, automation, reliability, and latency management.

The software suite includes AI-managed beamforming, AI-powered outdoor positioning, and advanced AI models capable of instant coverage prediction. These technologies work together to dynamically optimize network behavior based on traffic conditions, device activity, and service requirements.

Ericsson also announced new low-latency tools aimed specifically at supporting real-time AI and augmented reality experiences. Features such as Latency Prioritized Scheduler and Low Latency Mobility are designed to deliver significantly faster response times and more stable mobile connections for demanding applications.

According to the company, these capabilities could enable telecom providers to launch entirely new categories of premium AI connectivity services that were previously difficult to support at scale.

The portfolio expansion reflects a broader shift occurring across the telecommunications industry as operators prepare networks for the next phase of AI-driven internet usage. AI applications are not only increasing overall traffic volumes but also changing the way networks must prioritize speed, responsiveness, and reliability.

Ericsson says integrating AI directly into RAN infrastructure will also improve energy efficiency and reduce operational complexity for telecom providers, helping lower long-term infrastructure costs while improving user experiences.

Mårten Lerner said the company is embedding AI across its entire portfolio to deliver more intelligent and efficient network performance as consumer expectations evolve.

The new technology has already attracted attention from major telecom operators. VodafoneThree said it plans to integrate Ericsson’s AI-powered RAN infrastructure into its combined 5G network strategy, aiming to deliver faster, more reliable connectivity while reducing operational complexity.

Industry analysts believe AI-integrated telecom infrastructure could become one of the defining trends of the next phase of 5G evolution. As AI-powered devices and services become increasingly mainstream, network operators will likely face growing pressure to modernize infrastructure capable of supporting more intelligent, responsive, and data-intensive digital experiences.

Ericsson’s latest launch positions the company at the center of that transformation as telecom providers worldwide race to prepare their networks for the AI era.