Telecom Sovereignty Needs Its Own Framework
Much of the technology sovereignty debate centres on cloud infrastructure, AI and data. Which company controls the AI model? Who owns the datacentre? Which jurisdiction applies to information stored there? Who can access it, and could a foreign government somehow constrain or stop the service?
Data sovereignty concerns control over information and the rules governing it. Cloud sovereignty extends into the platforms processing that information, including operational control, software dependencies and the ability to change supplier if needed.
By contrast, telecoms sovereignty encompasses the networks and communications services connecting people, organisations and machines – as well as the datacentres for cloud and AI, plus critical infrastructure. It too involves scarce physical resources, international coordination and capabilities needed during emergencies.
There is a link to resilience here too. Is it possible to recover from problems, whether caused by government action, cyberattack or technology failure?
I find three additional concepts useful too: agency, independence and optionality. Can we decide and act of our own volition? How exposed are we to external interference? And can we change suppliers or architectures as circumstances evolve?
Telecoms dependencies go beyond those seen for cloud and data. Spectrum rights and satellite orbits require coordination beyond national borders. Mobile networks need precise timing to function, potentially dependent on satellite navigation systems. Subsea cables involve landing stations, transit countries and specialist repair capabilities. These concerns sit alongside familiar questions about network vendors, chips, software and cloud platforms.
Consider a UK organisation using a domestically hosted, notionally “sovereign” cloud service. Its connectivity might still depend on overseas network management and support, international transit or a foreign supplier's licence server or certificate platform. Even where normal traffic remains local, backup routing may behave differently. Sovereignty assessments should examine the customer traffic, the systems controlling its journey and what happens when something fails. That analysis could go as deep as understanding the supply of spare components, or how systems might be upgraded or network-software bugs patched.
Voice communications is a special case. Telephone numbers, caller identity and emergency calling have strong national characteristics, and in years past could have been considered absolutely sovereign, with a local supply-chain of switching equipment and user devices. Yet in the 2020s, a call may involve an international communications platform, cloud transcription or an AI assistant for translation or out-of-hours reception roles. Keeping a recording in Britain addresses only part of that chain. Control over call routing and service continuity matters too.
Now consider that a typical telecom network has a complex mesh of software for operational and business support (called OSS and BSS), doing everything from billing to network planning to scheduling maintenance visits, the dependencies are stark. These types of systems can take years to upgrade and transform – the idea of suddenly ripping out foreign-sourced software is completely unrealistic.
International inter-dependence is particularly important. Roaming, subsea connectivity and satellite services derive much of their value from crossing borders. Greater sovereignty may, counter-intuitively, require stronger agreements with overseas partners. There will be a need for diverse (and well-mapped) routes and interoperable systems. We also need to ask whose sovereignty we mean. Governments, telecom operators and enterprise customers have different priorities. A factory may want direct control of its private wireless network. A utility may need communications that remain available during a wider outage. Supporting a single “national champion” telco does not automatically advance all these objectives. While these are challenges, there are also telecoms opportunities around sovereignty. I group them into three categories:
Sovereign connectivity
Sovereign communications solutions
Fuller sovereign telecoms/cloud stacks.
A fibre provider offering assured routes between datacentres has a useful role, as does a specialist UC (unified communications) provider delivering locally-controlled voice services. Some large operators may also build cloud propositions, or entire datacentres for critical sovereign functions – although they will still have dependencies on foreign semiconductor and other hardware suppliers.
There are trade-offs here. Local operations and additional suppliers can improve control, but duplicated infrastructure adds cost. Sovereignty requirements should reflect the consequences of failure and the customer's actual needs. It is easy to over-specify, and ask for requirements inappropriate for more basic workloads or communications uses.
For the UK, I would like to see an explicit telecoms sovereignty framework developed alongside cloud initiatives. It should be developed with participation from specialist providers and enterprise networks as well as major operators. The framework should identify which overseas dependencies can reasonably be reduced, and which will need to be managed. It should assess legal, operational and technical control separately. Procurement processes should test for credible recovery and switching options.
That would align with policy goals, and also give buyers a practical basis for defining how much sovereignty they need and the likely costs involved.
