The shift from automation to autonomy is not a distant horizon for telecom – it is an active strategic choice operators are being forced to make now. Recently, a session moderated by Rishi Shukla, APAC Sales Head, Rakuten Symphony, brought together Faiq Khan, SVP & Head of Sales – EMEA at Rakuten Symphony, and Nitesh Singh, MD and Communications, Media and Technology Lead at Accenture, to examine what L4 and L5 autonomous networks will mean in practice: for the workforce, for governance, for the role of system integrators, and for the players in the industry who will emerge as winners.
Speakers:
Watch the full interview.
The fear that L4 and L5 autonomous networks will eliminate telecom roles wholesale is, the panel argued, a misconception worth dismantling. The more accurate framing is a shift in what people do, not a reduction in how many people are needed. Engineers will move from execution to governance, defining intent, architecting boundaries, and scripting the conditions under which AI operates.
The panel drew a parallel to the broader technology industry: just as software-defined networking did not eliminate network engineers but changed their skillset, autonomous networks will create demand for site reliability engineers and AI architects who define what the network is permitted to do.
The 80/20 principle applies: roughly 80 percent of routine, repeatable tasks will be handled autonomously; the complex 20 percent will remain firmly in human hands. Talent, the leaders agreed, will be the defining differentiator.
As AI takes on decision-making in networks that underpin national infrastructure, the governance question becomes unavoidable. The panel's position was clear: responsibility lies with the architects who define the system's intent and its blast radius. AI will do what it is designed to do, and no more, as long as those boundaries are well-defined.
The challenge is that this work is still in early stages. Regulatory frameworks, sovereignty considerations around data jurisdiction, and the question of who ultimately owns AI-driven decisions on critical national infrastructure are all active problems.
Rather than being designed out of relevance by autonomous networks, system integrators are positioned to become more central to the ecosystem. The panel was emphatic: the organizations that win will be those that architect automation into every layer of their stack from the ground up, with programmability and openness at the core. Mundane, routine tasks will be agent-defined. Complex integration, architectural judgment, and cross-ecosystem orchestration will require human expertise and that is where system integrators will play.
The panel's most pointed message was reserved for legacy players. Autonomy and open architecture create the conditions for genuine competitive differentiation – hyper-personalized services, AI-native network slicing, dynamic offers tailored to individual subscriber behavior in real time. But those same forces will commoditize or eliminate operators who remain locked in closed, box-based architectures without the data mediation layers needed to participate.
The speed at which AI and open architectures are evolving leaves little room for gradual adaptation. The winners will be those who build with competition and programmability at the heart of every architectural decision, starting now.
“There is no eradication of manpower in question. Instead of having software engineers, we’re going to have architects who are going to define the intent of AI systems. I see this as a huge opportunity.”
