Argyll launches UK sovereign AI cloud for organisations

SOFIAH NICHOLE SALIVIO

News Editor

Argyll Data Development has launched a sovereign AI inference cloud for UK organisations, designed to keep infrastructure and model control within UK jurisdiction.

The Dunoon-based company built the platform with SambaNova for organisations that want to run production AI workloads without relying on foreign-owned hyperscale cloud providers.

The launch comes as businesses and public sector bodies move AI systems from pilot projects into live operations, bringing greater scrutiny over where data is held, who controls the underlying systems, and how services meet regulatory requirements. In sectors such as defence, healthcare and finance, those questions have become more pressing because some workloads cannot be moved offshore.

Argyll says the platform combines UK-owned infrastructure with SambaNova hardware and software so that data, models and operations remain under UK control. It is intended to address concerns about reliance on overseas cloud groups for AI inference.

Sovereignty focus

At the centre of the service is SambaNova’s Reconfigurable Data Unit architecture, running the company’s SambaManaged system. The design can be deployed in existing UK data centres, with racks operating at about 10kW, in contrast to the higher power demands and cooling requirements often associated with GPU-based systems.

The cloud hosts open-source models including Minimax and can deliver speeds of up to 400 tokens per second within a UK-resident environment. It is designed for real-time AI applications ranging from customer operations to fraud detection.

Argyll has also structured the platform as a disaggregated system, allowing compute, storage and networking to be distributed across multiple UK locations while functioning as a single inference layer. The company says this offers resilience and flexibility for regulated and security-sensitive users.

Peter Griffiths outlined the company’s view of what constitutes sovereign AI infrastructure.

“Sovereignty in AI is not a label you can apply to a contract or a colocation agreement. It is a condition that has to be demonstrated – who is accountable, where the infrastructure sits, who controls the intelligence layer, and whether all of that aligns with the expectations of the society being served. Our platform satisfies those conditions. We are building the standard that others should be measured against,” said Peter Griffiths, Chairman of Argyll Data Development.

The launch reflects a wider debate in the UK over how AI services should be built and governed as adoption grows. Much of the market relies on large US cloud providers for computing and model access, but some organisations have raised concerns that dependence on overseas platforms could complicate compliance, procurement and public trust.

Energy use and operating costs have also become central issues as AI models are deployed at scale. Argyll and SambaNova are positioning their offer as an alternative to GPU-led systems, arguing that power consumption, cooling needs and ongoing infrastructure costs can become barriers when organisations move from testing to full production use.

Jude Sheeran, who leads SambaNova in Europe, the Middle East and Africa, said many users had not fully considered those trade-offs.

“As organisations scale AI, many are defaulting to GPU infrastructure without fully accounting for long-term cost, energy and operational complexity. Our work with Argyll provides an alternative, enabling high-performance AI inference that is more efficient, deployable and aligned with sovereignty requirements,” said Jude Sheeran, Managing Director for EMEA at SambaNova.

Argyll describes itself as a developer of renewable-powered infrastructure for AI in the UK. Its flagship project is the 184-acre Killellan AI Growth Zone in Argyll, where it plans to combine on-site wind, wave and solar generation with data-centre infrastructure.

That broader strategy links the company’s sovereign cloud pitch to domestic energy supply as well as data jurisdiction. For UK organisations deciding where to place sensitive AI workloads, Argyll is arguing that control over infrastructure, operations and location should sit together rather than be split across contracts and overseas cloud platforms.