Nokia advances autonomous network strategy with focus on trusted AI and scale

Rodrigo Brito is Vice President, Secure & Autonomous Networks at Nokia. He outlines how communication service providers can move beyond isolated use cases and scale autonomous networks to deliver measurable operational gains and support future AI-driven services.

TM Forum: What is the business case for Autonomous Networks? Why should telcos invest in higher levels of autonomy?

RB: The business case for autonomous networks is becoming clearer and more urgent. For telecommunication providers, the starting point is operational efficiency. Networks are becoming more complex, traffic patterns are changing faster, and customer expectations for reliability and performance continue to rise. At the same time, revenue growth remains under pressure, so operators need a fundamentally different operating model. Higher levels of autonomy help reduce manual effort, simplify operations, improve productivity and enable networks to respond at machine speed.

This becomes even more important as the AI super cycle accelerates. Traffic patterns are becoming more unpredictable, demand is shifting in real time, and networks increasingly need real-time control to support AI-driven applications and services. In this environment, autonomous networks become essential to deliver the adaptability, reliability and responsiveness that customers and enterprises will expect.

Fully autonomous networks will anticipate needs, solve problems before they arise and constantly optimize for peak performance. The goal is to move toward networks that are zero-wait, with instant order fulfilment; zero-touch, with automated delivery; and zero-trouble, with a flawless service experience. This is the kind of operating model telcos need if they are to manage growing complexity while improving customer experience.

But the opportunity is not only about cost reduction. Autonomous networks also create the foundation for new revenue models. When networks become programmable, intent-driven and able to act as one system across domains, telcos can launch and adapt services much faster. This is especially important as AI becomes a dominant workload on networks and enterprises need deterministic performance, security and flexibility.

To unlock that value, they need the right foundation: high-quality data, 360-degree observability, AI-powered intelligence, closed-loop automation and strong governance. Autonomy must also be trusted. In mission-critical telecom networks, operators need explainable AI, policy controls, auditability and the ability to supervise actions.

In short, telcos should invest in higher levels of autonomy because it addresses two strategic priorities at once: reducing the cost and complexity of running networks today, while creating the adaptability, reliability and programmability needed to monetize networks in the AI era.

TM Forum: How is Nokia helping your customers on the journey to autonomous networks?

RB: Nokia is helping customers move from isolated automation use cases toward autonomy as an operating model. We support that journey pragmatically: proving value in high-impact domains, scaling what works, and building the data, AI, orchestration and closed-loop automation capabilities needed for L4+ autonomy.

For example, stc Group in Saudi Arabia deployed Nokia’s AI-powered MantaRay AutoPilot for autonomous RAN operations, enabling 15,000 autonomous corrective actions per hour, 10% higher downlink throughput and 30% higher cell utilization. Nokia also delivered end-to-end service orchestration and closed-loop automation, resulting in a measured 40% to 50% productivity improvement.

We are also working with Netherlands-based KPN on its ambition to achieve Level 5 autonomy for 90% of services by 2030, showing how autonomy is becoming a long-term strategic transformation agenda across orchestration, assurance, data, AI operations and governance.

In North America, we helped a Tier 1 provider identify up to $30 million in savings through improved data management, reinforcing that trusted, high-quality data is the foundation for scaling AI and automation reliably across network domains.

Across these examples, Nokia is helping customers build autonomy step by step by combining AI-powered automation, service orchestration, closed-loop assurance, data intelligence and governance to deliver measurable gains in productivity, service quality and customer experience.

TM Forum: We’ve been talking about Autonomous Networks for years – what is the single biggest factor holding progress to higher autonomy?

RB: The single biggest factor holding progress back is consistent and reliable access to high-quality data. Autonomy depends on AI, and AI depends on data. If the data is fragmented, incomplete, inconsistent or difficult to access, even the most advanced AI models cannot deliver trusted outcomes.

Networks need to sense what is happening across domains, services and subscribers before they can reason, predict and act. That requires more than simply collecting large volumes of network data. It requires data that is curated, governed, observable and ready to use by AI models, automation workflows and autonomous agents.

This is one of the biggest challenges for telecommunication providers because many still operate with siloed systems, domain-specific tools and legacy data architectures. As a result, teams often spend too much time finding, preparing and reconciling data before they can even begin to apply AI. That slows innovation, limits trust in automation and makes it difficult to scale autonomy beyond isolated use cases.

Nokia’s view is that the industry needs to move from fragmented data lakes toward a more governed data mesh approach, where reusable data products can be created, trusted and consumed across applications and domains. With strong data management, data observability, clear data policies and governance, operators can create the single source of truth needed for AI-driven operations.

So, while there are several barriers to higher autonomy, including legacy systems, skills, integration complexity and organizational change, the most fundamental one is data. Without trusted, high-quality data, there is no trusted AI. And without trusted AI, operators will not be comfortable allowing networks to move from human-led automation toward higher levels of autonomous decision-making and action.

TM Forum: Everyone is talking about agentic AI, but what is sometimes overlooked in all the excitement?

RB: What is sometimes overlooked in the excitement around agentic AI is the importance of trust, governance and control. Telecom networks are mission critical and underpin the functioning of our society, so as we move toward more autonomous networks, we have to make sure that actions taken by AI agents are safe, accurate and explainable.

That means agentic AI cannot be treated as a black box. Telecommunication providers need to understand how decisions are made, what data and models were used, and why a specific action was recommended or taken. They also need the ability to trace, audit and observe outcomes across the full lifecycle of data, models, agents and network actions.

Governance is also essential because autonomous agents may have different objectives. One agent may optimize for energy efficiency, another for customer experience, another for capacity or cost. Without clear policies and guardrails, those objectives can create conflicts or unintended trade-offs. Telecommunication providers therefore need a framework that defines what agents are allowed to do, when human oversight is required, and how decisions are aligned with business, regulatory and operational priorities.

At Nokia, we call this approach Glassbox AI. It is about making AI transparent, governed and accountable, so operators can benefit from the speed and intelligence of agentic AI while maintaining the confidence needed in mission critical networks. In other words, the real breakthrough is not just creating agents that can act; it is creating agents that can be trusted to act.

TM Forum: Finally, what are you most excited about – what’s new for Nokia this year?

RB: What I am most excited about this year is how quickly the conversation is moving from automation to true autonomy. AI is becoming the dominant workload of the network economy, and that changes what networks need to be. They can no longer be static infrastructure. They need to become programmable, AI-native platforms that can anticipate demand, adapt in real time and operate at machine speed.

For Nokia, this is exactly where our autonomous networks strategy comes together. Our message this year is “autonomous networks built for the AI era” — complete, trusted and ready. Complete, because we have the foundation across domains, layers and technologies to help operators build future-ready autonomous networks. Trusted, because telecom networks are mission critical, and AI-driven autonomy must be transparent, governed and secure. Ready, because this is not just a future vision. We are already helping customers move toward Level 4 and beyond with real deployments, real use cases and measurable outcomes.

What is especially new and exciting is the focus on agentic AI and Glassbox autonomy. Agentic AI gives networks the ability to reason, recommend and act across complex operational scenarios. Glassbox autonomy makes that intelligence explainable, observable and governed, so operators can trust how data, models, agents and actions are being used. That combination is critical if the industry is going to scale autonomy safely in live networks.

What also makes this year exciting is that we are not doing this alone. Autonomous networks require an industry-wide ecosystem, and Nokia is working closely with customers, technology partners and industry bodies to turn the vision into practical, deployable solutions. At DTW Ignite, we are showcasing how these collaborations are coming to life through real use cases, demonstrations and proof points that show the industry moving from concept to execution.

This reinforces an important point: the AI era will require networks that are not only intelligent, but open, interoperable and trusted across a broader ecosystem. By working with partners across the industry, we can help accelerate the journey to autonomous networks and give operators the confidence that these capabilities are ready to scale in real-world environments.

So, I think this year is an important moment. We are showing how autonomous networks can help operators manage the complexity of the AI era, deliver deterministic performance for AI workloads, improve customer experience and create new monetization opportunities. The AI era will not just run on networks; it will be defined by them. And Nokia is ready to help telecommunication providers build those networks now.