TL;DR
At Cloud Native Telco Day during KubeCon + CloudNativeCon Europe 2026, Ranny Haiby and Philippe Ensarguet explained that Agentic AI in telecom builds on nearly a decade of cloudification and open source work. While the underlying technologies are advancing rapidly, successful deployment depends on addressing practical challenges, including trust and guardrails, OSS/BSS integration, data readiness, multi-vendor coordination, observability, and security.
At Cloud Native Telco Day, during KubeCon + CloudNativeCon Europe 2026, Ranny Haiby (Linux Foundation) and Philippe Ensarguet (Orange) explored the operational and organizational challenges involved in applying Agentic AI to telecom networks.
Watch their full keynote here.
Ranny opened by placing Agentic AI in the context of the telecom industry’s ongoing cloudification efforts. To help make sense of the growing number of technologies in the AI ecosystem, he presented a layered view of the stack, adapted from a framework shown by Jensen Huang, and highlighted the open source projects and initiatives that provide the building blocks for applying Agentic AI in networking.
His central message was that the industry is not starting from scratch. Operators and vendors have spent nearly a decade modernizing networks through NFV and work across the ecosystem. As he put it, “the basis for Agentic AI for networking is actually laying on top of what we all did with cloudification.”
Ranny also distinguished between two related areas of work: AI for networking, which uses agents to make networks more autonomous, and networking for AI, which focuses on building networks and telco networks suitable for running AI workloads.
Philippe shifted the discussion from architecture to operations. Drawing on his experience at Orange and conversations with industry peers, he summarized the core challenge succinctly: “Most of the time, the technology honestly is not the problem.”
The larger obstacles, he said, are operational and organizational. As networks move from intelligence to autonomy, telecom teams must establish trust and guardrails, integrate with long-standing OSS/BSS environments, prepare data spread across siloed systems, coordinate across multiple vendors and domains, and maintain observability, security, and accountability. Philippe captured a reality many operators know well: “We don’t have the API. We don’t have even the documentation.”
Together, the keynote underscored that applying Agentic AI to telecom networks builds on the industry’s existing cloud native foundation, but successful deployment depends on addressing the operational and organizational challenges involved in bringing these capabilities into production.
AI Disclosure
This post used artificial intelligence tools for research, structural assistance, or grammatical refinement. The final content was reviewed, edited, and validated by human contributors to to ensure accuracy and alignment with our community standards. We are committed to transparency in the use of generative technologies within the open source networking ecosystem.
