ETSI frames intent-driven orchestration to automate complex B2B streaming networks

Intent-driven orchestration marks a shift from manual configuration to outcome-based network management, crucial for maintaining high-quality video delivery across fragmented edge and cloud infrastructures. By abstracting resource-level complexity into 'intents,' streaming providers can automate fault recovery and traffic steering without vendor-specific scripts. This development bridges the gap between high-level service requirements and autonomous network agents, standardizing how AI-driven systems interpret and execute operational goals. Industry observers should watch for the release of the first normative draft in late 2026, which will signal the framework's readiness for multi-vendor interoperability trials.

The launch of DGS/ZSM-026 follows a series of advancements in the ETSI Industry Specification Group for Zero-touch network and Service Management (ZSM). According to ETSI technical reports from January 2026, the group has increasingly focused on the role of autonomous agents in managing the 'edge-to-cloud continuum.' This includes the publication of GR ZSM 020, which studied agent utilization in autonomous networks, providing a foundation for the current orchestration framework. These efforts aim to solve the scaling challenges posed by 5G and 6G architectures, where the density of network functions exceeds what human operators can manage.

Recent market activity highlights the practical demand for these standards. In June 2026 at Digital Transformation World (DTW), Huawei and China Telecom were recognized for deploying over 900 AI agents to manage radio network experiences, reporting a 20% reduction in poor user-experience incidents. Furthermore, the European ACROSS project successfully validated ZSM Proof of Concept #16 in early 2026, demonstrating multi-domain service management across geo-distributed testbeds in Greece and Spain. Per project documentation, these trials confirm that standardized intent-driven frameworks can maintain strict service level and security requirements across administrative boundaries.

ETSI's roadmap indicates that DGS/ZSM-026 is part of a broader cluster of new work items, including DGS/ZSM-028, which specifically focuses on managing Large and Small Language Models (LLMs/SLMs) in production network environments. By integrating intent-driven orchestration with generative AI management, ETSI is positioning its reference architecture to handle the next phase of agentic operations where natural language and intent serve as the primary interface for network control.