Explainer
AI sovereignty explained
What AI sovereignty means, the three layers it spans, and why it has become a board-level question.
By James Drayson
In short
AI sovereignty is the degree to which you control the AI you depend on, the data it processes, the infrastructure it runs on, and the model itself. It applies at two levels: national (a country's control over its AI) and organisational (an enterprise's control over the AI it uses).
National vs organisational sovereignty
- National AI sovereignty: A country's ability to develop and run AI on its own infrastructure, in its own languages and laws, without dependence on foreign providers.
- Organisational AI sovereignty: An enterprise's ability to own and control the AI behind its operations, keeping data, IP, and decisions inside its perimeter.
The three layers of sovereignty
- Data sovereignty: Your data stays in the jurisdiction and perimeter you choose, and is never absorbed into a third party's systems.
- Infrastructure sovereignty: The compute runs where you control it, on-prem, in your cloud tenant, or a sovereign cloud.
- Model sovereignty: You hold the weights and IP and decide the update cadence, the layer most providers leave out.
Why it's a board-level issue
AI is becoming core infrastructure for regulated organisations. When a critical capability depends on a model you don't own, hosted abroad, changeable under you, that is concentration risk, regulatory exposure, and loss of a strategic asset, all board-level concerns. Sovereignty reframes AI from a tool you rent into an asset you own and govern.
What this looks like with Locai
Sovereignty stops being a contractual promise when the machine is standing in your own building. Here is what that looks like in practice.
Locai Labs builds Locai One, an on-prem AI appliance. It is one machine that arrives with everything already in it: the hardware, our open-weight Locai Juno models, and Locai OS, the operating system that serves the models and handles users, access and monitoring. You plug it into a mains socket and your network, and your team is working in about 15 minutes. No cloud account, no per-token bill, and nothing leaving the building.
The reason a data-centre-class model fits in a box on your floor is SPACE, our compression algorithm. Instead of asking how much of a model can be cut while keeping it broadly similar, SPACE asks what the model needs to be good at, preserves the subnetworks behind those capabilities and strips back the rest. The result is a smaller specialist rather than a shrunken generalist, tuned to the exact hardware it ships on.
Locai One starts at £29,950 for a team, and Locai One Pro at £49,950 for an organisation, bought once and owned outright. Both run air-cooled on standard mains power and work fully air-gapped. Any compatible open-weight model runs alongside Juno, and if you need a model trained on your own proprietary data we can post-train one and deploy it on the same machine.
Frequently asked questions
What is AI sovereignty?
It is control over the AI you rely on, the data, the infrastructure, and the model, so the capability is genuinely yours rather than dependent on a third party.
What's the difference between data and model sovereignty?
Data sovereignty keeps your data in your jurisdiction; model sovereignty means you own the weights and control how the model changes. You need both for full sovereignty.
Why does AI sovereignty matter for business?
Because depending on a model you don't own creates regulatory exposure, concentration risk, and loss of a strategic asset, increasingly a board-level concern.
How do you achieve AI sovereignty?
Own the model weights and run them inside your perimeter under your jurisdiction, on-prem, air-gapped, or in a sovereign cloud, with training on your own data.
Book a sovereign AI briefing
A 30-minute session on owning your model: deployment options, the data path, and a clear cost range for your use case.