Seattle, October 6: Microsoft has formalised a Sovereign AI framework, built with Nvidia, aimed at governments and regulated industries that need strict control over data.

The framework rests on four principles: control, choice, flexibility and resilience. It offers two deployment tiers, Sovereign Public Cloud and Sovereign Private Cloud, sitting above Microsoft Foundry, Foundry Local and Azure Local.

A joint white paper with Nvidia bundles Confidential Computing, NVIDIA AI Enterprise, Nemotron models and RTX PRO infrastructure as the reference stack. It is pitched as protection for sensitive data, models and AI workloads during processing.

"Sovereign AI is the design, deployment and operation of AI workloads under defined controls," wrote Didier Ongena, Microsoft's Public Sector and Sovereignty Leader.

The move puts Microsoft against sovereign-cloud pitches from Oracle, AWS and European hyperscalers. Europe is flagged as a key region, though no launch customer or contract value was named.

The framework covers connected, hybrid and fully disconnected modes. Microsoft says durable AI strategies need enough control to use AI confidently and enough choice to keep adopting better technology.