AI-RAN Won’t Be a Reset: Why Operators Are Viewing AI as Next Layer of Open RAN

The telecom industry has spent the better part of a decade transforming the radio access network (RAN). Operators have virtualized network functions, embraced cloud-native architectures, deployed Kubernetes-based infrastructure, and invested heavily in Open RAN ecosystems. Now, as the industry intensely discusses AI-RAN, many operators are asking: Does this require another major architectural overhaul?

According to Paul Miller – CTO, Intelligent Systems Software and Services, Aptiv – the answer is no.

In a recent TelecomTV interview, Miller argued that the transition from cloud-native RAN to AI-native RAN should not be viewed as another disruptive migration. Instead, he asserted, AI-RAN represents the next phase in the evolution of Open RAN and virtualized RAN, building on infrastructure that operators already deployed rather than replacing it.

That message aligns closely with the priorities currently shaping telecom network strategy: operational efficiency, AI-driven automation, energy savings, and preparation for 6G. Industry organizations including the O-RAN Alliance and the AI-RAN Alliance have been framing AI as an extension of open, cloud-native networks rather than a separate architectural track.

AI-RAN Builds on Existing Cloud-Native Investments

AI-RAN is not a new architecture, Miller explained. The fundamental cloud-native design introduced through Open RAN remains intact.

It was a massive undertaking for the industry to move from 4G appliance-based systems to virtualized and cloud-native 5G infrastructure. Operators moved from vertically integrated hardware platforms to distributed, multi-vendor architectures based on commercial servers, virtualization software, containers, and Kubernetes orchestration. Those foundational elements remain relevant in the AI era.

This perspective is consistent with the direction set by the O-RAN Alliance, which positions virtualization, openness, and intelligence as complementary pillars of future RAN architectures. Recent O-RAN priorities emphasize AI-enabled network management, RAN Intelligent Controllers (RICs), Service Management and Orchestration (SMO), and AI/ML lifecycle management as critical enablers of future deployments.

For operators, the implication is clear: AI-RAN should extend prior investments rather than invalidate them.

The First Killer App May Be Cost Reduction

Although much of the industry discussion surrounding AI focuses on new revenue opportunities, Miller pointed toward a more immediate and practical use case: reducing operating costs.

Specifically, he highlighted AI-enabled radio optimization techniques such as dynamic beamforming and intelligent power management. Given that radio infrastructure accounts for a significant share of network operating expenses, even modest efficiency improvements can generate meaningful savings.

The focus on economics reflects conclusions from analyst firms including Dell’Oro Group, Omdia, and Analysys Mason. While their forecasts differ in detail, they consistently observe that operators remain under pressure to improve network economics as traffic growth continues to outpace revenue growth. In that environment, AI projects that save money often receive stronger support than initiatives built around speculative future revenues.

Data Becomes a Strategic Asset

Miller described AI-RAN as a continuous feedback loop built around “sense, think, act, and optimize.” In his view, “The ‘sense’ really implies that you’re collecting data from the edge systems and allowing that to adjust your application and model,” he said. “And then you drive the change back to that system through an optimized layer that provides continuous lifecycle management and updates of those AI inference applications.”

Historically, network telemetry existed primarily to support monitoring, troubleshooting, and service assurance. In an AI-RAN environment, telemetry also becomes the fuel that drives model training, inference, optimization, and automation.

Industry organizations also emphasize this requirement. TM Forum’s work on AI-native operations, described in “Driving AI Automation across IT and Networks,” focuses heavily on governance, lifecycle management, trust, security, agentic AI, and production-scale AI systems.

For operators, this means AI-RAN adoption likely requires investment beyond AI infrastructure, also including analytics platforms, data pipelines, orchestration tools, and operational processes that can support continuous model improvement.

The Road to Monetization Remains Important

That does not mean operators have abandoned hopes of generating new revenue from AI.

AI-RAN could eventually support new monetization opportunities, noted Miller, particularly as operators continue searching for ways to expand beyond traditional connectivity services.

This remains a strategic discussion in the telecom industry. According to GSMA Intelligence research  presented at Mobile World Congress 2026, operators are exploring AI infrastructure services, enterprise AI platforms, edge AI, GPU-as-a-Service offerings, and differentiated connectivity products.

Examples are already emerging. Deutsche Telekom made AI the centerpiece of its MWC 2026 strategy, showcasing AI-powered services across telecom networks, enterprise cloud, digital identity, smart homes, and emerging 6G infrastructure. Under the theme “Magenta AI at Scale. Human at Heart,” the company highlighted how AI is being integrated into networks, products, and customer experiences.

Still, most operators appear to view these opportunities as a second phase of the AI-RAN journey. Today’s focus remains firmly on efficiency and automation.

AI-RAN and the Path Toward 6G

Miller suggested that future 6G networks will almost certainly be built on cloud-native and virtualized foundations. If that assumption proves correct, today’s Open RAN and cloud-native investments become critical preparation for tomorrow’s AI-native network architectures.

Industry discussions increasingly support that position. O-RAN Alliance leaders have repeatedly identified virtualization, cloud-native deployment models, and AI-enabled control systems as foundational capabilities for future network evolution.

For operators that have not yet fully embraced virtualization and cloud-native operations, the implication is significant. The path to AI-RAN may ultimately be less about deploying AI models and more about completing the foundational transformation required to support them.

The Path Forward for AI-RAN

If a single message is emerging from recent operator, vendor, standards-body, and analyst discussions, it is that AI-RAN’s initial success will be measured less by futuristic applications and more by practical operational outcomes.

Energy savings, automation, network optimization, autonomous operations, and improved utilization consistently rank among the industry’s highest priorities. Organizations ranging from the O-RAN Alliance and AI-RAN Alliance to TM Forum, NVIDIA, and GSMA Intelligence all point toward these benefits as the most immediate drivers of adoption.

The most important takeaway from Miller’s interview, therefore, is not that AI-RAN is coming. That much is already clear. The more important message is that operators do not need to start over.

The telecom industry’s years-long investment in Open RAN, virtualization, Kubernetes, cloud-native infrastructure, and distributed operations appears increasingly likely to serve as the foundation for AI-RAN adoption. AI is the next layer of intelligence that sits on top of them. 

For operators, the immediate opportunity lies in energy savings, operational efficiency, automation, and improved network performance. Longer term, AI-RAN may unlock new monetization models and play a central role in the industry’s transition toward 6G.

That combination of near-term business value and long-term strategic relevance helps explain why AI-RAN has become one of the most closely watched developments in telecom. More importantly, it suggests that the industry’s next transformation may feel less like a revolution and more like the natural continuation of a journey that is already underway. The future of telecom isn’t just cloud-native or AI-driven — it is the seamless combination of both.

AI-RAN builds on the cloud-native foundations operators have already established. Discover how Wind River helps service providers deploy scalable, intelligent infrastructure for Open RAN, automation, and future 6G networks by exploring how to enable Edge AI for telco RAN, core, IT, and enterprise.

Monetize 5G and accelerate cloud-native network transformation for vRAN, Open RAN, core, and IT with a distributed cloud Edge AI management platform with ease of deployment and operations at scale. Explore Wind River’s Cloud Platform for telecommunications.