What happened
SoftBank Corp. announced a strategic shift at MWC Barcelona 2026, evolving from a telecom carrier to an AI infrastructure provider. Its "Telco AI Cloud" vision transforms the network into an AI-native platform, orchestrating distributed AI workloads across edge and cloud via AI-RAN-based Multi-access Edge Computing (MEC). This delivers real-time inference for "Physical AI" applications like robotics, offloading GPU compute. SoftBank demonstrated optimised robot connectivity with Ericsson, deployed its 'AITRAS' platform for industrial edge AI with Mitsubishi Heavy Industries, and open-sourced the AITRAS Orchestrator's Dynamic Scoring Framework.
Why it matters
Real-time AI inference at the network edge redefines operational capabilities for physical AI systems. Robotics engineers and platform architects gain immediate decision-making for resource-constrained devices, as SoftBank's Telco AI Cloud embeds AI models directly into the network via AI-RAN-based MEC. This offloads GPU compute, supporting complex behaviours beyond onboard hardware limits and ensuring stable, low-latency connectivity. Procurement teams require network infrastructure with AI-native capabilities, integrating edge compute for real-time physical AI deployments. This follows Nokia's recent expansion of AI network partnerships.
Subscribe for Weekly Updates
Stay ahead with our weekly AI and tech briefings, delivered every Tuesday.
