Microsoft Expands Control Over Data, AI Systems and Cloud Infrastructure

Microsoft is continuing to expand its digital sovereignty capabilities through Microsoft Sovereign Cloud, bringing together public cloud services, customer-operated private infrastructure and partner technology environments. The new capabilities are aimed at organizations managing sensitive data, critical information systems and business processes subject to strict regulatory requirements, including government agencies, financial institutions, healthcare providers, energy companies and other critical infrastructure operators.

The growing adoption of advanced cloud services and artificial intelligence is raising questions that go far beyond the physical location of stored data. Organizations increasingly need to understand who can access their systems, who controls the underlying infrastructure, how sensitive digital assets are protected, and how heavily their services depend on individual technology providers. Another important consideration is whether essential services can continue operating during regulatory or geopolitical changes, security incidents, or disruptions to network connectivity.

Microsoft introduced its Sovereign Cloud platform last year to provide customers with greater flexibility in how they manage infrastructure and data. Since then, development has covered public cloud services, infrastructure operated directly by customers, and solutions delivered through technology partners with expertise in local markets. The latest expansion introduces additional technical and operational controls, greater transparency into system access and the ability to run more applications and AI workloads within locally controlled environments.

Douglas Phillips, President and Chief Technology Officer of Microsoft Specialized Clouds, has emphasized that digital sovereignty requirements are increasingly tied to organizations' ability to adopt advanced technologies without surrendering control over their most important systems. This is particularly relevant to government institutions, critical infrastructure operators, and organizations in regulated industries, where technological development must be balanced against security, operational, and legal obligations.

Different Infrastructure Models for Different Workloads

Microsoft Sovereign Cloud is built around the principle that the appropriate level of control should be determined for each application or workload individually, taking into account data sensitivity, operational requirements, technology dependencies and business continuity needs.

This approach allows organizations to combine different infrastructure environments without requiring every application to follow the same deployment model. A public-facing digital service may benefit from the scalability and continuous innovation of the public cloud, while a system supporting essential operations may need to run locally and remain available even when connectivity is lost.

For AI systems processing particularly sensitive information, additional requirements may include keeping models, data, and computational processing entirely within infrastructure directly controlled by the customer.

Microsoft has therefore organized its offering into three interconnected components: Sovereign Public Cloud, Sovereign Private Cloud and Sovereign Partner Ecosystem.

Sovereign Public Cloud is designed for organizations seeking to use Azure infrastructure and services while maintaining additional controls over data, system access and compliance with regional requirements. Sovereign Private Cloud extends cloud capabilities to customer-operated infrastructure, including environments that can function both online and offline.

The third component, Sovereign Partner Ecosystem, encompasses independently operated infrastructure, specialized technology providers, and local partners with knowledge of national regulatory frameworks and industry-specific requirements.

Organizations can use all three models simultaneously, adjusting the degree of control and operational independence according to the importance and requirements of individual systems. This approach also allows them to maintain consistent governance and oversight across different parts of their IT infrastructure.

Greater Transparency Into Public Cloud Access

Within Sovereign Public Cloud, Microsoft is introducing additional capabilities for monitoring compliance with digital sovereignty requirements, including data residency, confidentiality, infrastructure operator access, security policy enforcement, and business continuity.

One of the new capabilities is the Sovereign Control Panel, which will enter public preview at Microsoft Ignite. The tool provides a unified management experience within the Azure portal, enabling organizations to monitor and manage their digital sovereignty controls.

It allows users to review existing safeguards, identify gaps and risks, and take corrective action. Beyond security and regulatory compliance, the platform also supports business continuity planning by analyzing application dependencies and assessing whether workloads can be moved between different operating environments when necessary.

Access control is further strengthened through existing mechanisms such as Just-in-time Access and Customer Lockbox, which enable customers to manage access to their environments, particularly when intervention by Microsoft technical personnel is required.

For customers in the European Union and European Free Trade Association (EFTA) countries, Microsoft is gradually introducing Data Guardian. The capability provides oversight by a trusted EU-based partner for qualifying remote access by Microsoft engineers to production systems covered by the program.

Microsoft is also rolling out Production Touch Ledger, which generates records of access to production systems and the actions performed during those sessions. These records are designed to make unauthorized modifications detectable, providing organizations with an additional means of reviewing activities carried out within their environments.

Both capabilities are particularly relevant to organizations required to demonstrate regulatory compliance, monitor privileged infrastructure access, and document activities performed by external technology providers.

Encryption Keys Can Remain Outside Microsoft Infrastructure

Another important development concerns cryptographic key management and the protection of sensitive data. Azure Key Vault Managed Hardware Security Module (HSM) already provides encryption key protection through dedicated, single-tenant hardware security modules.

Microsoft is now extending this approach through external key management with Azure Managed HSM, which has become generally available.

Organizations can retain their root encryption keys on hardware they own and operate, either on their own premises or through a trusted third party. This allows physical control over a critical component of cryptographic infrastructure to remain separate from Microsoft's data centers.

The capability is particularly relevant to systems where internal security policies or regulatory requirements demand direct organizational control over the keys used to protect sensitive information.

Microsoft is also expanding its confidential computing capabilities, which protect data while it is being processed inside hardware-protected and verifiable trusted execution environments.

Azure Confidential Clean Rooms is currently in limited preview and enables approved participants from multiple organizations to perform joint analyses of sensitive data while limiting exposure of the underlying information.

Such an approach can be used when several organizations need to collaborate on data analysis but are unwilling or unable to provide other participants with broader access to their original datasets.

By combining encryption key control, protected execution environments and access oversight, Microsoft aims to give organizations additional technical safeguards that can support broader digital sovereignty strategies aligned with their legal, security and business requirements.

Critical Systems Must Remain Operational Without Internet Connectivity

Digital sovereignty is increasingly associated with operational resilience, particularly for infrastructure that must continue functioning during security incidents, network outages, or changes in geopolitical circumstances.

Microsoft is therefore developing new guidance and architectural patterns for business continuity in sovereign infrastructure environments. These resources are intended to help organizations translate resilience requirements into practical technical architectures that can be planned, deployed and tested in advance.

For many applications, the public cloud provides resilience through geographically distributed data centers and the ability to deploy systems across multiple locations. However, for certain critical business processes, dependence on network connectivity can itself represent an unacceptable operational risk.

Sovereign Private Cloud therefore supports both connected and disconnected operations. The ability to transition workloads from connected to disconnected environments is entering public preview.

This capability gives organizations greater flexibility to maintain essential services when connectivity conditions or operational requirements change.

It is particularly relevant to government systems, industrial facilities, energy infrastructure, and other organizations requiring a high degree of operational independence. Nevertheless, the ability to operate locally does not eliminate the need for established recovery procedures, data governance and regular business continuity testing.

Azure Local Expands Local Application Management

Greater operational independence extends beyond the infrastructure on which applications run. Organizations also need locally available databases, development tools, AI models and productivity systems on which those applications depend.

Microsoft Sovereign Private Cloud is built on Azure Local, which extends Microsoft cloud capabilities to infrastructure operated and controlled directly by customers.

SQL Server on Azure Local is already generally available for both connected and disconnected deployments. DocumentDB and Blob Storage on Azure Local are entering public preview, expanding the range of technologies available for storing and managing business-critical data within customer-controlled environments.

These developments allow organizations to keep more of their data infrastructure within their own operational environments instead of requiring individual application components to rely on remote public cloud services.

The changes also extend to software development tools. GitHub Enterprise Local will become generally available at Microsoft Ignite as a validated GitHub Enterprise Server appliance designed for Azure Local.

This will enable organizations to retain source code repositories, development artifacts, and related information within locally managed infrastructure.

Developers will also be able to use GitHub Copilot with locally hosted models for inference, allowing AI-generated responses and suggestions to be produced by models running within the customer's own environment.

This approach brings AI-assisted development closer to the source code and underlying data, reducing the need for development processes to depend entirely on external infrastructure.

AI Models and Agents Are Moving Closer to Local Infrastructure

Microsoft is also expanding local artificial intelligence capabilities through Foundry Local.

At Microsoft Ignite, Foundry Local is expected to become generally available for connected and disconnected operations on Azure Local. Its supported model catalog will also expand to include Microsoft models, partner-developed models, open-weight models, and open-source alternatives.

This gives organizations more choice in selecting AI models and deployment approaches according to their security, regulatory, performance, and business requirements.

A particular focus is the development of local inference and agentic retrieval capabilities. These systems allow users to query internal business information using natural language and receive responses grounded in data available within their organization's local environment.

The approach enables generative AI and agent-based functionality without necessarily transferring sensitive business data to remote processing environments.

Integration with Microsoft 365 Local, available in preview to approved customers, extends these capabilities to locally operated Microsoft SharePoint and Exchange servers.

Organizations can therefore connect locally deployed AI models to business documents, communications systems and other internal content while retaining control over the infrastructure used to process that information.

These developments also expand the scope for AI deployment in industries where regulatory or security requirements restrict the use of external services. However, the actual level of control and protection will depend on deployment architecture, model selection, and data governance policies.

Confidential Virtual Machines and Additional Hardware Protection

Alongside local AI and data services, Microsoft has made confidential virtual machines on Azure Local generally available.

This capability extends confidential computing to infrastructure managed directly by customers. Protection relies on hardware-encrypted memory and hardware-based isolation, reducing the exposure of sensitive data during processing.

Additional assurance comes from workload attestation, a process used to verify the integrity and security state of an execution environment before granting access to sensitive data or cryptographic keys.

These mechanisms are particularly relevant to regulated, security-sensitive, and mission-critical applications where data must be protected not only during storage or transmission but also while it is being processed.

Azure Local is also gaining support for multitenancy. Organizations will be able to separate computing resources, governance policies, and administrative functions across multiple tenant environments while sharing the same underlying physical infrastructure.

This capability is important for large enterprises with multiple business units, organizations requiring separate security domains, and providers of infrastructure and communications services.

Local operators and service providers will be able to develop distinct sovereign cloud environments for different customers on shared Azure Local infrastructure, applying separate governance models and security controls.

As a result, local infrastructure will increasingly support almost the entire application lifecycle, from data storage and management to software development, artificial intelligence, productivity tools and day-to-day operations.

National Clouds and Local Partners Become Part of the Sovereignty Model

Beyond its own infrastructure and software solutions, Microsoft is further developing the Sovereign Partner Ecosystem, bringing together national cloud providers, software companies, infrastructure manufacturers and specialized technology integrators.

This approach reflects the differences in regulatory and operational requirements between countries and industries. For some organizations, keeping data physically within a particular jurisdiction is not sufficient. They also require specific ownership structures, infrastructure operating models, or control over administrative procedures.

National Partner Clouds, including Bleu in France, illustrate this approach by providing independently operated environments for customers with particular ownership and operational control requirements.

Microsoft's software partners extend Azure Local capabilities across cybersecurity, identity management, observability, backup and recovery, migration, data management, networking and applications.

Partners specializing in digital sovereignty combine Microsoft technologies with local regulatory knowledge and implementation expertise. Other partners contribute to the development of infrastructure platforms ranging from data centers to edge computing environments.

The partner ecosystem includes Accenture, AMD, Armada, Avanade, base-it, Bleu, Cirrus Data, CloudCasa by Catalogic, Cronos Groep, cTelecoms, DataON, Delos Cloud, Dell Technologies, F5, Felo, Fortinet, HPE, Inspark, Intel, Mattermost, Mistral, NVIDIA, Qualidev, Qumulo, RiverMeadow, Rubrik, Supermicro, Tata Consultancy Services and Veeam.

By involving infrastructure, security and software providers, Microsoft is allowing customers to combine its technologies with products and services from other vendors.

For national and regional cloud service providers, this model creates opportunities to develop specialized offerings for the public sector, regulated industries and organizations seeking locally managed infrastructure.

However, such arrangements do not necessarily deliver complete technological independence. That depends on infrastructure ownership, licensing relationships, technical support arrangements, application architecture, and the ability to migrate to alternative platforms.

Digital Sovereignty Becomes a Central Issue in AI Development

The expansion of artificial intelligence is introducing additional challenges for digital infrastructure control. AI models and agents are increasingly accessing sensitive information, participating in business processes, and connecting to information systems responsible for important organizational functions.

Unlike traditional applications, AI agents can perform more complex tasks, use multiple data sources, and take actions within information systems. Security requirements therefore extend beyond data protection to include defining permitted activities, monitoring access, and retaining the ability to restrict or terminate the actions of individual agents.

Organizations need clear control over where AI systems operate, which people and processes manage them, what permissions they receive, and how business continuity will be maintained when operating conditions change.

Microsoft is consequently placing greater emphasis on combining public cloud services, locally managed infrastructure and partner environments as part of its digital sovereignty strategy. Organizations can establish different levels of control for individual systems based on their business importance and security requirements, rather than applying a single infrastructure model to every application.

Local AI model execution, encryption key control, administrative access oversight, and the ability to continue operating without connectivity to external data centers are becoming increasingly important elements of this approach.

The latest Microsoft Sovereign Cloud capabilities illustrate the direction in which the market for regulated cloud services is evolving. The focus is shifting from the physical location of data toward broader control over infrastructure, operations, software dependencies, and artificial intelligence.

For organizations operating critical systems, however, it remains essential to distinguish between data sovereignty, operational independence and full technological autonomy. Local data processing and privately controlled infrastructure can provide greater control, but they do not automatically eliminate dependence on software vendors, proprietary technologies or support arrangements.

The ability to assess these dependencies, implement appropriate levels of control, and maintain critical operations during disruptions will increasingly influence how government institutions and enterprises select infrastructure for their data, applications, and AI systems.