Ericsson has unveiled AI in RAN, as the vendor seeks to embed AI tools within its Radio Access Network (RAN) platforms.

Launched on June 11, it will enable Ericsson to bring telco-grade AI models directly into basebands and radios to boost 5G networks without the need for new hardware.

According to Ericsson, its AI in RAN is a software subscription that utilizes AI to boost efficiency and performance, and to help make energy savings.

The vendor claims this will bolster 5G networks for communications service providers (CSPs) and support the shift to AI-native RAN without requiring additional hardware.

The company states that AI in RAN introduces telco-grade AI models designed to run in real time within the RAN, continuous learning software powered by scalable, high-quality data, and Agentic AI support to enable advanced RAN automation and network operations.

“Ericsson is redefining what’s possible in mobile networks by bringing powerful AI capabilities to service providers,” says Mårten Lerner, head of networks strategy and product management at Ericsson. “With AI in RAN software, we are taking a major step toward AI-native networks, alongside the AI-ready radios we unveiled in February.”

Ericsson notes that its telco-grade AI models have been designed for ultra-low-latency inference at the microsecond level.

“Service providers need networks that are high-performing, secure, and efficient in the AI era,” Lerner adds. “With AI in RAN, Ericsson is bringing AI into networks to elevate 5G performance and efficiency through energy-efficient AI inference at scale.”

AI in RAN works with Ericsson 5G Advanced across both purpose-built and cloud RAN platforms to enable new AI-driven services, while AI in RAN is designed to use the right AI model in the right part of the radio network.

Ericsson said it's able to do this through Ericsson Silicon for energy-efficient AI inference in radios and the latest generation of RAN compute, while the portability of cloud RAN software enables AI capabilities to be deployed across partner platforms.

The first features are already available, going live earlier this quarter, while further enhancements to the service will come later in the year.

Providing some extra details on these features, Ericsson said these include an AI-native scheduler for link adaptation, AI-powered macro positioning, AI-managed beamforming, AI-powered multi-layer coordination, performance management event schema files, and augmented observability for AI in RAN.

The vendor has carried out more than 15 deployments and trials worldwide of its AI in RAN software subscription, claiming that it delivers up to 20 percent higher downlink throughput and up to 10 percent better spectral efficiency.

Ericsson also claims that it supports up to 2x more high-traffic users, 90–95 percent coverage prediction accuracy, and achieves up to 5x greater user-positioning precision.

Ericsson's AI in RAN has been used by several telcos, including SoftBank Corp, Bell, SK Telecom, and Rogers.