Will Sovereign AI Become a New Revenue Stream for Operators?

July 2026

Telecoms & Connectivity

Ardit Ballhysa

Senior Research Analyst

In the pursuit of generating additional revenue streams, mobile network operators have been searching for ways to diversify beyond traditional connectivity services. This was particularly exacerbated shortly after operators had built out their 5G infrastructure, as the copious amounts of investment did not translate into like-for-like revenue growth, and many operators are still failing to generate significant revenue from 5G.

The onset and growth of sovereign AI - the development and deployment of AI created and maintained within a country’s or region’s borders - has emerged as a potential opportunity for operators to expand beyond their traditional connectivity role. Across Europe, governments and enterprises are increasingly prioritising AI infrastructure that is hosted, operated, and governed within their borders because of concerns surrounding data sovereignty, regulatory compliance, and dependence on foreign cloud providers. Because operators own many of the assets required to support distributed AI infrastructure, including fibre networks, edge computing sites, and datacentres, sovereign AI has emerged as a clear opportunity; positioning operators well to extend beyond being just traditional connectivity providers and into being regional AI infrastructure providers. Being positioned well, however, is not enough, as operators will be pitted against hyperscalers, who dominate the AI infrastructure market.

From Connectivity to Infrastructure

Unlike traditional cloud computing, many AI applications increasingly require computational resources to be located closer to where data is generated, ie the edge. AI inference – the process of running trained AI models to produce real-time outputs – is becoming essential for applications requiring low latency, such as autonomous manufacturing, industrial automation, intelligent transport systems, and healthcare monitoring. Processing these workloads at the edge reduces latency, minimises bandwidth consumption, and improves overall system responsiveness.

This suits operators well, as they maintain thousands of distributed network sites connected through high-capacity fibre infrastructure. Additionally, many have already invested in edge computing platforms as part of their broader 5G strategies. While these investments have yet to generate significant returns, sovereign AI could provide a commercial use case capable of unlocking their value. Furthermore, AI is starting to become entrenched within operators’ networks through AI-Radio Access Network (RAN), which combines AI workloads with traditional mobile network infrastructure. As operators begin to deploy graphical processing units within edge infrastructure to support AI-RAN, edge infrastructure could become a shared computing platform capable of supporting both operator services and commercial AI workloads.

Working with Hyperscalers, Not Against Them

Despite these opportunities, becoming an AI infrastructure provider will be a significant strategic shift for operators, and one that presents significant challenges. Companies such as Alibaba, Amazon, Microsoft, and Google have spent decades building highly mature cloud platforms (Alibaba Cloud, Amazon Web Services (AWS), Microsoft Azure, and Google Cloud, respectively), extensive developer ecosystems, and global AI infrastructure. These providers already dominate enterprise cloud computing, and continue to invest billions of dollars into AI-specific infrastructure. Combined, these four hyperscalers spent $416 billion on infrastructure in 2025 alone; a scale that is not feasible for operators. As such, competing directly with hyperscalers would yield no positive results for operators. The former possess far superior software expertise, cloud-native platforms, scale, and developer communities.

Although hyperscalers have introduced sovereign AI offerings, these are being delivered through partnerships with operators; whereby hyperscalers contribute AI platforms and software, and operators provide the domestic infrastructure, connectivity, and regulatory compliance needed to satisfy sovereignty requirements.

A New Role Rather than a New Identity

As such, operators will find greater success by positioning themselves as providers of sovereign, distributed AI infrastructure, and partners for hyperscalers’ sovereign solutions. Private 5G networks are an example of how this strategy could evolve. Enterprises deploying private networks increasingly require AI-powered analytics, automation, and computer vision applications that benefit from processing data at the edge. By integrating AI inference with edge computing and private 5G infrastructure, operators could offer fully managed platforms that combine connectivity, computing, and AI services. This would allow operators to move further up the value chain while leveraging assets they already own.

The emergence of sovereign AI also complements broader European efforts to strengthen digital independence. As governments continue investing in domestic AI capability, operators could play an increasingly important role in providing the distributed infrastructure necessary to support national AI ecosystems. For instance, Deutsche Telekom (DT), Orange, Telefónica, TIM, and Vodafone announced during Mobile World Congress 2026 that they had built a federated European Edge Continuum explicitly framed around digital sovereignty; creating a European-owned digital infrastructure. Earlier in 2026, in partnership with NVIDIA, DT launched a €1 billion Industrial AI Cloud targeting enterprise, government, and defence customers.

This trend of operators partnering with each other, and with hyperscalers, will expand throughout other countries and regions globally; positioning operators as key players of enabling sovereign AI. Subsequently, operators will be able to monetise their infrastructure beyond traditional connectivity; although the degree to which they can achieve this depends largely on emerging AI use cases and adoption of the technology.