Microsoft has expanded its Sovereign Cloud portfolio with capabilities designed to support governance, productivity, and large artificial intelligence models in fully disconnected operational environments.
The release centers on three main additions: Azure Local disconnected operations, Microsoft 365 Local disconnected, and enhanced Foundry Local support for large multimodal models. These features are part of the broader Sovereign Private Cloud stack, which unifies infrastructure, productivity services, and AI model execution across connected, hybrid, and wholly offline deployment modes.
Azure Local disconnected operations are now generally available and enable organizations to run mission-critical infrastructure under consistent Azure governance and policy controls without dependence on public cloud connectivity. This addresses continuity requirements in regulated, classified, or intentionally isolated environments where network dependencies can introduce risk or violate compliance mandates.
Alongside this, Microsoft 365 Local disconnected brings core productivity workloads — including Exchange Server, SharePoint Server, and Skype for Business Server — into sovereign operational boundaries. These workloads operate entirely within the customer’s controlled environment and are supported through at least 2035, ensuring that essential collaboration capabilities remain operational even when disconnected from external cloud services.
The update to Foundry Local broadens support for larger, multimodal AI models running on modern hardware within sovereign private clouds. Leveraging partner-provided infrastructure such as NVIDIA GPUs, organizations can now perform local inferencing and API-driven AI tasks entirely within their own environments, retaining full control over data and model execution without requiring connectivity to public cloud endpoints.
Operationally, these changes reflect growing enterprise demand for digital sovereignty — not just data residency but end-to-end control over compute, governance, and compliance in scenarios where external connectivity is either restricted or unacceptable. For industries such as government, defense, financial services, and healthcare, the ability to standardize governance, maintain productivity, and execute advanced AI models in air-gapped settings aligns with strict regulatory requirements and risk profiles.