Open RAN, Cloud RAN and AI Converge: 5 Insights Operators Need Now
Introduction
On July 30, 2026, RCR Wireless News hosted a webinar featuring Téral Research founder Stéphane Téral, Nokia Head of AI-RAN and Cloud RAN Aji Ed, and TELUS Director of RAN Strategy Sushil Rawat. The discussion framed AI-RAN not as a standalone hype cycle but as the natural convergence of Open RAN openness, Cloud RAN virtualization, and AI-driven intelligence. For Open RAN operators, this signals a pragmatic evolution: Open interfaces are already delivering the flexibility needed for AI applications, while real deployments show measurable capacity gains without requiring wholesale hardware rip-and-replace.
The message is clear: operators should focus on outcomes—spectral efficiency, automation safety, and brownfield modernization—rather than vendor-specific branding. Here are five key insights that matter most for technical and business decision-makers.
1. Open RAN Is the Programmable Foundation, Not a Failed Experiment
Téral pushed back on narratives that Open RAN has underdelivered. Instead, he positioned it as the enabler of choice: “No, it did not [fail]. It gives you flexibility. That means you pick and choose whoever as a vendor or supplier you want in your network.” Open interfaces create a programmable base layer for cloud-native functions and AI apps, evolving from earlier self-organizing network (SON) efforts dating back to 3GPP Release 8.
For operators already on the Open RAN path, this means AI-RAN dApps and rApps can plug into existing RIC frameworks without starting from scratch. The convergence accelerates because disaggregation removes vendor lock-in, allowing best-of-breed AI algorithms to optimize radio resources dynamically.
2. Nokia’s AI-Native Platform Delivers Immediate Capacity Gains on Open RAN Specs
Nokia’s anyRAN software combined with NVIDIA’s Aerial AI-RAN platform offers three flexible paths: an accelerated plug-in for existing AirScale deployments, standalone AI-RAN nodes, and cloud-native implementations on COTS servers. All adhere to Open RAN specifications and share a common software architecture.
The payoff is tangible. AI-driven radio algorithms have already demonstrated more than 20% spectral-efficiency gains, with a clear roadmap to 50% by 2027 and more than 100% by 2028. Ed summarized the ambition: “We bring twice the network capacity, twice the spectral efficiency compared to what we have today, and we are not going to stop there.”
Critically, deployment is workload-driven rather than processor-driven. Operators can choose GPU acceleration where economics justify it, avoiding a one-size-fits-all silicon mandate.
3. TELUS Shows Pragmatic Brownfield Open RAN Scaling with AI in Mind
TELUS began its multi-vendor Open RAN program in late 2023 during a hardware refresh. Today it represents 25% of the network, targeting 40% by year-end 2026, 50% by end of 2027, and 100% by 2029. Interoperability is treated as an operational reality, not just a standards checkbox.
On AI-RAN specifically, Rawat emphasized outcomes over taxonomy: “It’s basically driven by outcome, right? It really doesn’t matter what you call it.” TELUS does not anticipate needing GPUs at cell sites within the next 12 months, but centralized accelerated compute is already relevant for model training, digital twins, and anomaly detection.
This timeline aligns AI investments with existing Open RAN rollout milestones, reducing risk for operators following similar brownfield strategies.
4. Production-Grade AI Requires Guardrails Beyond Lab Demos
Moving AI from demonstration to safe, scaled production remains the harder challenge. Rawat highlighted identity and access management, conflict controls, guardrails, and integration with change-management processes as essential.
“You can build a use case. You can demonstrate it in lab, quick and easy. Taking it to the production, scaling it for day-to-day operation. This is a very important aspect of it.”
Operators must treat AI models with the same rigor as network configuration changes—version control, rollback capabilities, and observability are non-negotiable. Open RAN’s disaggregated architecture actually helps here by allowing AI functions to be isolated and monitored independently.
5. The Long-Term Vision Is Shared Infrastructure and Edge AI Services
Téral described today’s phase as primarily “AI for RAN”—applying machine learning to boost performance and efficiency. The next horizon involves shared infrastructure running both RAN and AI workloads, followed by edge-based AI services for enterprises and new revenue streams.
Despite lingering questions around business cases, the direction is irreversible: “This is where we are going. You know, there is no way back.” Open RAN and Cloud RAN supply the disaggregated, virtualized substrate; AI supplies the intelligence layer that improves economics and reliability.
Conclusion
The July 30 webinar underscored that AI-RAN maturity depends less on flashy announcements and more on measurable integration with Open RAN foundations. Operators who treat Open RAN as a strategic platform rather than a checkbox will be best positioned to capture capacity gains, automate safely, and prepare for edge AI opportunities. The convergence is happening now—through software evolution, not revolution.