Home Archives for November 18, 2026
Cloud Services
Sovereign AI and Data: What Australian Organisations Need to Know
-
Luke de Merindol
- 8 min read
Summary
- Sovereign AI is the ability to develop, deploy and run AI while keeping control of the infrastructure, data and governance behind it, including who operates the technology and which laws apply.
- Data residency is where data is physically stored. Data sovereignty is the legal authority and control over that data. Keeping data in Australia supports sovereignty but doesn’t guarantee it on its own.
- AI apps can send prompts, outputs and logs offshore even when the main database stays local, so businesses need to map the full data lifecycle.
- Australia’s National AI Plan (December 2025) makes domestic AI capability, infrastructure investment and responsible adoption national priorities.
- The main AI data security risks are sensitive data exposure, unauthorised access, data integrity problems, third-party dependencies and governance gaps.
- A sovereign AI foundation needs data governance, the right infrastructure, security controls, a clear picture of provider responsibilities, and plans for scaling.
- Macquarie Cloud Services supports sovereign AI through its Managed Data Platform (built on Microsoft Fabric) and Launch® AI, a managed sovereign AI platform that keeps data, prompts and logs in Australia.
- Macquarie was the first company to have all its data centres and cloud services Certified Strategic under the Australian Government’s Hosting Certification Framework. It holds ISO/IEC 27001:2022 and PCI DSS v4.0.1 certification and has SOC 1 and SOC 2 reports.
- Sovereignty doesn’t replace security or compliance. Organisations still need to meet the Privacy Act 1988, the Australian Privacy Principles and the rules for their industry.
Artificial intelligence is changing how businesses work, from automating everyday tasks to uncovering insights hidden in large volumes of data. But as AI becomes more deeply embedded in business operations, an important question is emerging: who controls the data powering it?
Every AI application relies on data. Whether it’s a customer service assistant, financial forecasting tool or intelligent automation platform, that information needs to be stored, processed and protected somewhere.
For businesses handling sensitive information, knowing where that happens, who can access it and which laws apply is becoming increasingly important.
This is where sovereign AI and data sovereignty come into focus.
As Australia expands its AI capabilities, organisations need to balance innovation with security, governance and control. Understanding what sovereignty means in practice is an important first step.
What Is Sovereign AI?
Sovereign AI refers to the ability to develop, deploy and operate artificial intelligence systems while maintaining control over the infrastructure, data and governance supporting them.
Rather than focusing solely on where an AI model is hosted, sovereignty considers who operates the technology, how information is processed and which jurisdictions may have authority over it.
A sovereign AI environment may involve several important elements:
- Data control: Understanding how information is collected, stored, processed and accessed.
- Australian infrastructure: Hosting AI workloads within Australia to support data residency and operational requirements.
- Security and governance: Establishing appropriate access controls, monitoring and oversight.
- Operational independence: Reducing unnecessary reliance on external providers and maintaining flexibility over AI systems.
- Regulatory alignment: Supporting applicable Australian privacy, security and industry obligations.
The objective isn’t necessarily to develop every AI model or technology locally. It’s to ensure organisations have appropriate control over how AI operates and how their information is managed.
What's the Difference Between Data Sovereignty and Data Residency?
Although the terms are often used interchangeably, data sovereignty and data residency describe different concepts.
Data residency refers to the physical location where information is stored or processed. For example, a business may choose to host its data in an Australian data centre to meet operational or contractual requirements.
Data sovereignty, on the other hand, concerns the legal authority and control governing that information.
Keeping data in Australia can support sovereignty objectives, but physical location alone doesn’t guarantee complete control. The ownership of the infrastructure, the provider’s legal obligations and the way information is accessed or transferred can also affect sovereignty.
For AI workloads, the distinction becomes even more important.
An AI application may store its primary database locally while sending prompts or other information to an external model for processing. Depending on the service, logs, outputs or operational data may also be handled in different locations.
Understanding the complete data lifecycle helps businesses identify where information travels and which controls are needed to protect it.
Why Does Sovereign AI Matter in Australia?
AI adoption is creating new opportunities across financial services, healthcare, education, government and other industries. At the same time, it introduces questions about data protection, accountability and operational resilience.
Australia’s National AI Plan, released in December 2025, identifies domestic AI capability, infrastructure investment and responsible adoption as priorities for the country’s AI development.1
The Australian Government has also recognised the importance of secure, onshore data centre infrastructure in supporting sovereign AI capabilities.2
These developments reflect a growing need to ensure AI innovation is supported by appropriate infrastructure and governance.
For businesses managing sensitive workloads, sovereign AI can help address several practical concerns.
It can provide greater visibility into where information is processed, support data residency requirements and give organisations more control over how AI services are deployed and managed.
However, sovereignty is not a substitute for security or regulatory compliance. Organisations still need to assess their legal obligations, security controls and the responsibilities of their technology providers.
What Are the Data Security Risks of AI?
AI introduces new ways for information to be processed, analysed and shared. While these capabilities can improve productivity, they can also create security risks when sensitive data is introduced into systems without appropriate controls.
The Australian Signals Directorate’s guidance on AI data security highlights the importance of protecting information throughout the AI lifecycle, from data collection and preparation to deployment and ongoing operation.3
Some of the key considerations include:
- Sensitive data exposure: Confidential information may be disclosed through prompts, outputs, logs or connected systems.
- Unauthorised access: Weak identity controls can expose AI applications and the information they process.
- Data integrity: Inaccurate, manipulated or compromised information can affect the reliability of AI outputs.
- Third-party dependencies: External AI platforms may introduce additional security, privacy and jurisdictional considerations.
- Governance gaps: Limited visibility into AI usage can make it difficult to monitor risks and enforce organisational policies.
These challenges highlight why AI security needs to begin with the underlying data and infrastructure.
Before deploying AI applications, businesses should understand which information will be used, how it will be protected and whether existing security controls remain appropriate.
How Can Businesses Build a Sovereign AI Foundation?
Successful AI adoption requires more than selecting a model or deploying a new application.
The quality of the underlying data, the infrastructure supporting workloads and the governance surrounding AI usage all influence how effectively the technology can operate.
A strong foundation begins with understanding where business information resides and how it moves between systems. Fragmented data environments can make it difficult to maintain consistent access controls, monitor usage and establish reliable AI outputs.
From there, organisations can focus on several priorities:
- Establish data governance: Define how information is classified, accessed, retained and used by AI systems.
- Assess infrastructure requirements: Determine where AI workloads should run based on security, performance, residency and operational needs.
- Implement security controls: Apply identity management, encryption, monitoring and appropriate network protections.
- Understand provider responsibilities: Review how external platforms handle prompts, outputs, logs and other sensitive information.
- Plan for scalability: Consider how infrastructure, costs and governance will evolve as AI moves from experimentation into production.
The aim is to make security and governance part of the AI architecture from the beginning, rather than adding controls after applications have already been deployed.
How Does Macquarie Cloud Services Support Sovereign AI?
For Macquarie Cloud Services, sovereign AI is about helping businesses move beyond experimentation while maintaining control over the data and infrastructure supporting their AI workloads.
Through its Data and AI portfolio, Macquarie Cloud Services provides managed data and sovereign AI capabilities designed to help customers establish governed data foundations and deploy AI into production.4
Its Managed Data Platform uses Microsoft Fabric to help businesses connect information across systems, improve data governance and prepare trusted data for analytics and AI applications.
For organisations ready to deploy generative AI, Launch® AI provides a managed sovereign AI platform delivered from Australian private cloud infrastructure.5
The service is designed to support enterprise AI workloads while keeping data, prompts and operational logs within Australia. It also offers a capacity-based pricing model and a vendor-neutral approach to model integration, helping businesses manage costs and reduce unnecessary platform dependency.
By combining Australian-based infrastructure with managed cloud and data expertise, Macquarie Cloud Services helps customers address the practical challenges of scaling AI securely.
This approach is particularly relevant for organisations managing sensitive information, complex technology environments or requirements around data residency and operational control.
Sovereignty You Can Check
Any provider can call itself sovereign. The real test is the evidence behind the claim. Here’s what stands behind ours, across Macquarie Technology Group.
- Certified Strategic: Macquarie was the first company to have all its data centres and cloud services Certified Strategic under the Australian Government’s Hosting Certification Framework.6,7 Strategic is the framework’s highest level, with strict ownership and control conditions.
- Trusted by government: Macquarie Government helps secure 42% of the Federal Government. It has supported government agencies for more than 20 years.8
- Cleared, onshore people: More than 270 Australian-based engineers and architects hold NV1 or higher security clearances.9 Our cloud infrastructure, data centres, operations and staff all sit within Australia.
- Independently audited: Macquarie Cloud Services holds ISO/IEC 27001:2022 certification across its private cloud, public cloud, colocation and hybrid hosting services.10 It is also PCI DSS v4.0.1 compliant and holds SOC 1 and SOC 2 reports.11,12
- Defence ready: Our Tier III data centres in Sydney and Canberra are DISP certified and SCEC zoned for sensitive government workloads.7
For AI workloads, that means your data, prompts and logs stay with an Australian-owned provider. Australian people, cleared to government standards, run the environment every day.
Rather than treating sovereignty as a separate consideration, Macquarie Cloud Services incorporates it into the infrastructure and governance supporting AI adoption.
Is Sovereign AI Enough to Ensure Security and Compliance?
Keeping AI workloads within Australia can support data sovereignty objectives, but it doesn’t automatically make an environment secure or compliant.
A locally hosted AI application can still be exposed through weak access controls, vulnerable software, poor data governance or incorrectly configured infrastructure.
Likewise, Australian data residency does not automatically satisfy every requirement under the Privacy Act 1988, the Australian Privacy Principles or applicable industry-specific regulations.
Security and compliance depend on how information is managed throughout its lifecycle.
This includes understanding which data AI systems can access, how models are deployed, who can interact with them and what happens when information is no longer required.
Australia’s National AI Centre also recommends that organisations assess AI-specific risks, maintain appropriate governance and apply security controls throughout the development and operation of AI systems.13
Sovereignty therefore works best as part of a broader approach that brings together data protection, cloud security, responsible AI practices and operational resilience.
Building AI with Confidence Starts with Data Control
AI presents significant opportunities to improve productivity, automate processes and generate new business insights. But realising those benefits requires confidence in the information and infrastructure supporting the technology.
Sovereign AI gives businesses an approach to maintaining greater control over where AI workloads operate, how data is handled and which security and governance measures apply.
For organisations moving AI from pilot projects into everyday operations, these considerations are becoming increasingly important.
Macquarie Cloud Services supports this journey through Australian sovereign cloud infrastructure, managed data services and Launch® AI, helping customers establish the foundations needed to deploy AI securely and at scale.
Because successful AI adoption isn’t simply about what the technology can do. It’s also about knowing that the data behind it is being managed responsibly.
Take Control of Your AI and Data Strategy
Whether a business is preparing its data for AI or looking to move existing AI applications into production, the right infrastructure and governance can make a meaningful difference.
Macquarie Cloud Services helps organisations build governed data foundations and deploy sovereign AI solutions designed around security, control and operational requirements.
Talk to a Macquarie Cloud Services Data and AI expert about building a secure, scalable AI environment that keeps business data under control.
Resources
¹ Australian Government, Department of Industry, Science and Resources, National AI Plan, December 2025.
² Australian Government, Department of the Prime Minister and Cabinet, Getting It Right: Building AI Infrastructure That Works for Australia, September 2026.
³ Australian Signals Directorate, AI Data Security, May 2025.
⁴ Macquarie Cloud Services, Data and AI Solutions.
⁵ Macquarie Cloud Services, Launch® AI – Australian Sovereign AI Platform.
⁶ TelcoNews Australia, Macquarie Technology first to achieve DTA’s full data centre certification, December 2023.
⁷ Macquarie Cloud Services, Our Data Centres.
⁸ Macquarie Government, Secure Azure and Cyber Security for Government.
⁹ Macquarie Government, Secure DISP Enclave.
¹⁰ Macquarie Cloud Services, ISO/IEC 27001:2022 Certificate.
¹¹ Macquarie Cloud Services, PCI DSS v4.0.1 Certificate of Compliance.
¹² Macquarie Cloud Services, Compliance.
¹³ National AI Centre, Guidance for AI Adoption: Implementation Guidance.
Luke de Merindol
Contact us
Talk to an expert.
Let’s talk about where you are and what better looks like for your business.
- 1800 004 943
- Level 15, 2 Market Street Sydney, NSW, Australia







