The ITU FG-AINN workshop in Beijing brought together industry leaders including Huawei, ZTE, and China Mobile to discuss the standardization of AI-native network architectures. The sessions focused on integrating compute and communication resources, the evolution of agentic AI in telecommunications, and the development of a token-based economy for network infrastructure.
The shift toward AI-native architectures marks a transition from treating intelligence as an overlay to making it a structural requirement of the telecommunications stack. For the streaming ecosystem, this deep integration of compute and communication suggests that future networks will evolve from simple traffic pipes into active value hubs capable of on-demand token delivery. As autonomous agents begin to operate and optimize infrastructure, the industry must adapt to a token-based economy where unit costs for computation collapse while aggregate demand scales. Watch for the standardization of interoperability protocols between multi-agent systems and the deployment of the Wireless World Model for network automation.
The ITU's Focus Group on AI for Networks (FG-AINN) has been building momentum toward formal recommendations since its establishment in 2024. Huawei has positioned itself as a leading contributor to the group's work on integrating compute and communication resources into network architecture. At MWC Barcelona 2025, Huawei unveiled its Net5.5G vision emphasizing AI-native network capabilities for the 2030 era, framing the architecture as a prerequisite for handling the compute demands of autonomous agents and immersive applications. ZTE has similarly committed resources to the standardization effort, with the company presenting its AI-native RAN framework at the ITU workshop series in early 2025, proposing that base stations evolve into distributed compute nodes capable of running inference workloads locally. China Mobile, as the world's largest mobile operator by subscribers, has signaled operational readiness by deploying AI-driven network optimization across its 5G standalone core in 12 provinces during 2025, reducing manual intervention in traffic steering by 40 percent.
On the regulatory and business side, the ITU's standardization timeline carries weight because IMT-2030 specifications will influence national spectrum allocation and infrastructure procurement decisions across member states. The ITU Radiocommunication Sector published its IMT-2030 framework in November 2024, identifying AI integration as one of six key usage scenarios for next-generation mobile systems. Huawei has aligned its commercial roadmap accordingly, with the company announcing in March 2025 that it would invest $10 billion over three years in AI-native network R&D, targeting both carrier-grade deployments and enterprise edge scenarios. The token-based economy concept discussed at the Beijing workshop mirrors broader industry moves toward usage-based pricing for network compute, a model that Ericsson explored in its 2025 Mobility Report as a potential revenue stream for operators managing AI workloads at the edge.
From a technical standpoint, the AI-native network specifications intersect with competing approaches to network automation that streaming and content delivery stakeholders should monitor. The 3GPP has been developing its own AI/ML integration framework for Release 20, with the organization completing its first study items on AI-driven radio resource management in mid-2025, creating a parallel standardization track that operators will need to reconcile with ITU recommendations. Meanwhile, the Open RAN ecosystem offers a different path to network intelligence, with the O-RAN Alliance publishing its AI/ML workflow specification in January 2025 that defines how third-party AI applications can interface with radio access network components. For streaming infrastructure planners, the convergence of these standards will determine whether AI-native networks expose programmable APIs for content delivery optimization or remain opaque to application-layer services.
The ITU FG-AINN workshop in Beijing established new AI-native network specifications for IMT-2030 infrastructure. By integrating compute and communication resources directly into telecommunications architecture, this shift moves intelligence from an overlay to a structural requirement, enabling autonomous agents and a token-based economy to manage surging network demands and optimize future streaming.
The specifications aim to integrate compute and communication resources directly into telecommunications architecture, moving intelligence from an overlay to a structural requirement for IMT-2030 networks.
The workshop proposed a token-based economy where tokens serve as the fundamental unit of AI computation and value circulation, allowing for usage-based pricing as network compute demands scale.
ZTE introduced the NetNexus platform to support agentic AI evolution in telecom operations and has proposed an AI-native RAN framework where base stations function as distributed compute nodes for local inference.
China Mobile has deployed AI-driven network optimization across its 5G standalone core in 12 provinces, successfully reducing manual intervention in traffic steering by 40 percent.
Add StreamingMeme as a preferred source on Google to see more of our streaming news at the top of your Search results.
Add as preferred source
