Explainer
Data residency in AI
Keeping data, and the model that processes it, inside a defined jurisdiction, and why a model you own makes that guarantee simple.
By James Drayson
In short
Data residency in AI means keeping your data, and the inference that processes it, within a defined geographic or legal jurisdiction. The most direct way to guarantee it is to run a model you own inside your own perimeter, rather than sending data to a third-party model hosted elsewhere.
Why residency is hard with external APIs
Even when a provider offers regional hosting, your data is still processed by a model and organisation outside your control, often with sub-processors and terms that can change. For regulated data, that uncertainty is the problem.
A sovereign model removes the uncertainty: if the weights are yours and the model runs inside your perimeter, the data and its processing simply never leave the jurisdiction you choose.
Locai's residency options
- On-premise: Runs entirely in your own data centre.
- Your cloud tenant: Deployed inside your existing cloud region and account.
- Air-gapped: Fully isolated for the most sensitive data.
- UK sovereign cloud: Locai-operated, UK-resident hosting where you still own the model.
What this looks like with Locai
Compliance is far easier to evidence when the machine, the model and the logs are all things you own, not things a vendor describes in a whitepaper.
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
Does regional cloud hosting guarantee residency?
It helps, but the model and provider are still external and governed by their terms. Owning the model and running it in your perimeter gives a stronger, simpler guarantee.
How does residency relate to GDPR?
Residency is a key part of demonstrating GDPR compliance for personal data; keeping inference onshore avoids many cross-border transfer questions.
Can I choose where my Locai model runs?
Yes, on-prem, in your cloud tenant, air-gapped, or in Locai's UK sovereign cloud, whatever meets your residency requirements.
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.