London-based Valarian, which allows companies to use US cloud providers for AI workloads but retain control of their data, raised a $50M Series A led by NEA
Max Buchan started advocating for infrastructure sovereignty when, as he puts it, “globalization and Davos were still cool.”
Context & Ripple Effects
Valarian’s financing arrives amid a broader buildout of London’s AI infrastructure and governance ecosystem. Related coverage includes NexGen Cloud’s GPU-as-a-service funding, Geordie AI’s governance-platform round, and major London expansions by AI developers.
The company is positioned at a specific tension in AI adoption: customers want access to large US cloud platforms for compute while preserving control over their data. NEA has also backed cloud-data optimization company Granica, underscoring investor attention to the infrastructure layer around AI workloads.
First-order effects
- Valarian gains $50 million to develop and sell its infrastructure-sovereignty offering, while enterprises using its platform can pursue AI workloads on US cloud providers without ceding the same degree of data control.
- The round gives Valarian a stronger position with customers for whom data control is a gating requirement for cloud-based AI adoption.
Second-order effects
- Cloud infrastructure providers and AI platform vendors may face greater customer demand for architectures that separate access to compute from control over data, creating room for intermediary infrastructure layers such as Valarian.
- London’s AI stack becomes more differentiated: GPU access, governance tools, and sovereignty-focused infrastructure address adjacent constraints that can determine whether enterprises deploy AI at scale.
Third-order effects
- If this model proves repeatable, AI infrastructure could become less of a binary choice between operating privately and relying directly on hyperscale clouds; control and compliance layers may become a durable part of the cloud-AI market.
- The pattern points toward sovereignty and governance becoming product-level buying criteria for AI infrastructure, not merely legal or procurement conditions—though adoption will depend on whether such layers can preserve the performance and economics customers expect from major cloud platforms.
The trend: AI infrastructure is evolving from a pure compute race into a market for controlled, governed access to cloud capacity, with data sovereignty emerging as a core enterprise requirement.