“… the story of AI’s industrialisation is intimately tied to Cloud infrastructure …”¹

The United States (US) has long recognised the imperative to apply innovative commercial technologies like artificial intelligence (AI) to national security.² Transformative developments of AI will impact many areas of society, including Defence.³ For example the Maven Smart System is a powerful AI platform developed by the US Department of Defense, designed to identify personnel and equipment on the battlefield using machine learning algorithms. It integrates various data sources, including satellite imagery and geolocation data, into a single interface, enhancing decision advantage in combat situations.⁴

The perpetual threat-counter threat cycle of military conflict demands that any advantage is sought, analysed, and if found advantageous, exploited to the full. The ongoing information revolution has offered new avenues to be pursued. Innovations in chipset design coupled with ever improving AI training are now starting to yield tangible advantages, even in the early light of the quantum computing dawn. Consistent with the opening quote, organisations are increasingly turning to secure Cloud as their infrastructure foundation and to accelerate AI adoption.

This article will outline why AUKUS needs a Cloud by describing the necessity beyond a unified communications platform, to where Industry and AUKUS partners can collaborate faster on the development and introduction of AI capabilities. Critically, the AI outcomes the three nations seek to achieve will influence the design of workloads and choice of AI models, which in turn will indicate which Cloud is best suited to rapidly reach the potential of AUKUS.⁵

Why does AUKUS need a Cloud? 

The US Government’s Cloud Program is positioning the US to meet these information and data challenges.⁶ The goal of the US Department of Defense’s (DoD) Joint Warfighting Cloud Capability (JWCC) program is to provide Cloud computing services “at the speed of mission, at all classification levels, from headquarters to the tactical edge.”⁷ The US’s partnership with Australia and the UK under the AUKUS agreement arguably necessitates a minor adjustment to this statement by adding that it will be achieved “with key allies and industry partners.” Providing access to an AUKUS Cloud will improve productivity across the three nations through a unified communication platform.⁸ It will provide access for deployed users to retrieve and process data rapidly and securely in the battlespace, as well as enabling access for AUKUS users to enterprise systems from anywhere. Critically, it will also enable collaboration between AUKUS partners on the development and introduction of AI capabilities faster. This is important as not only will AI enable almost all technologies within AUKUS Pillar One and Pillar Two, but AI capabilities will also offer deterrence when indicators spike and deliver decision advantage when crises arise.⁹

AI and Cloud computing have a symbiotic relationship. Enterprise leaders should strive to understand how the rapidly expanding field of AI continues to develop a relationship with cloud computing technology that spurs ever-greater innovation. Cloud services offer the AUKUS partnership the ability to securely collaborate, deploying AI algorithms and models at scale, deliver better outcomes faster, expand the quality and quantity of training data, expand access to talent across the AUKUS enterprise, and improve interoperability through common understandings of legal and ethical issues. Exploiting AI technologies brings enormous potential to the AUKUS enterprise that can be further advanced through closer cooperation with all levels of Government, international allies, and industry.¹⁰ Government and industry will both need to keep their relationship agile if they are going to successfully introduce and update software and systems that respond to a rapidly changing world.¹¹ 

The AUKUS Cloud will enable each country to collaborate with one another and their respective Defence Industrial Base (DIB) partners. Improved collaboration with DIB partners will accelerate and improve the development, acquisition, delivery and support of Defence capability. Converging industry-specific applications with the capabilities of the Cloud enables information technology teams to deploy and manage specialised software, and helping ensure it aligns with capability outcomes.

Importantly for small-to-medium enterprises, Cloud platforms are crucial elements of the AI ecosystem, providing Cloud computing services that allow developers to perform computationally intensive AI work.¹²

The opportunity for sovereign small-to-medium enterprises on Cloud isn’t lost on governments. As the CIO for the Australian Department of Defence recently stated: “The real issue here is the overlying services that deliver a military effect on top of the hyperscaler. The hyperscaler is just a fabric. It’s the services that are on top of that, and that’s where our sovereign capabilities can support us, in developing those military capabilities that sit on top of the hyperscaler.”¹³ Indeed, recent Pentagon spending indicates an increasingly different class of defence contractors is emerging: a combination of large Cloud Service Providers and hundreds of smaller startup companies, including: Anduril Industries, Shield AI, HawkEye 360, Skydio, Rebellion Defense, and Epirus, among many others.¹⁴ For small-to-medium companies challenged by the Defence procurement environment and low public investment, AUKUS Cloud has the potential to better enable innovative technology exposure to a wider customer base, and ultimately placed in the hands of the warfighter faster.

While the US Department of Defense and its industry partners are working on leveraging AI, the pace of AI adoption across the AUKUS partnership is not moving at the speed or coordination demanded by the rapidly changing global security environment. In the US, positive efforts already underway to speed this adoption through the office of Chief Digital and Artificial Intelligence Officer and Task Force Lima,¹⁵ have been established to identify the best use cases and problem sets where Defense can work with industry and rapidly apply AI.¹⁶ A similar methodology of prioritisation and coordination could be applied across AUKUS, thereby reducing duplication and saving resources.

To accelerate AI outcomes, AUKUS will need to coordinate and prioritise AI objectives. As a recent RUSI/SCSP¹⁷ article suggests, working closely with industry partners, AUKUS could use AI to achieve significant advantages as part of a broad strategy, including but not limited to:

• Absorb and effectively use ever-increasing volumes of data at the operational and strategic levels. Properly programmed and supervised machines, enabled by a data-sharing regime, are well suited to augment human cognitive bandwidth, generate better situational awareness, and informing plans. The UK Chief of Defence Intelligence, Adrian Bird, emphasised the importance of “… a point of coherence for all of Defence’s future intelligence, surveillance and reconnaissance (ISR) related programmes. Presently, our ability to collect data risks overmatching the ability of our analysts to analyse and exploit it at the speed we need to. We expect that machine learning, and then artificial intelligence, will speed that process.”¹⁸ He suggests that to achieve this: properly programmed machines provide ‘… a data analytics ecosystem, which can ingest and store data from a variety of sources and to apply innovative data analytics and visualisation techniques. It should also have the necessary processing and algorithmic capability to support the automation of the intelligence requirements and processing cycles and of our collection management capabilities.’¹⁹

• Develop a faster, higher-quality decision-making cycle. AI can accelerate military decision-making processes to enhance decision advantage. This creates more time for planning, and improved ability to react to adversary operations faster. For instance, AI-based decision support systems are being developed to display, synthesize, and analyse data, and in some cases make recommendations or predictions to aid human decision-making in conflict.²⁰ Different parts of the military’s decision-making processes can be automated using AI technologies to extract observations from data streams, automatically build models of the terrain, generate predictions of future events and develop courses-of-action.

• Automate routine actions and processes to augment human decision-making at scale. Some tasks for tactical operations, operations centres, planning, maintenance and sustainment require limited decision-making skills, but consume considerable time and attention. Automating a Common Operating Picture (COP), where multiple reporting is instantaneously projected to a combined overlay in real-time, will save valuable time and effort. Automating these processes at scale would release human bandwidth for more complex tasks, potentially reduce the amount of personnel required for certain staff components of expeditionary operations, accelerate the planning process, and augment the quality and range of human decision-making. A large range of examples currently used elsewhere include healthcare, personnel management, finance, preventative maintenance, and logistics all of which have military applications.

What is the way ahead?

Cloud adoption is not solely an information technology decision nor a single Government decision, but a pan-AUKUS priority.²¹ AUKUS partners will need to partner with Cloud Service Providers to obtain access to the required hyperscale Cloud services. These services include computing, storage, database, network, big data storage and analytics, artificial intelligence, machine learning, robotics and synthetics.²² Given the current disparate levels of understanding and adoption of Cloud capabilities in each of the AUKUS partner countries, this will require leadership and prioritisation from senior leaders to fully realise the opportunity.

There are a range of factors for AUKUS partners to consider when determining how to progress to accessing Cloud.  This is about solving the right problems with the right technology. To begin with, including Cloud collaboration within extant national Defence Cloud strategies, or ideally by releasing separate AUKUS Cloud, Data, and AI strategies, would greatly assist Industry’s understanding of the challenges AUKUS programs face and prioritisation of AI development.

This is critically important as the AI outcomes AUKUS seeks will influence the design of workloads and choice of AI models, which in turn will indicate which Cloud is best suited.

While all hyperscalers offer powerful Cloud services, each catering to specific use cases, what AUKUS requires is the ability to migrate successfully while supporting hybrid Cloud deployments, integrate with a common productivity ecosystem, and access an array of pioneering AI and machine learning services. Critically, the AUKUS Cloud should have a focus on compliance and security features, and is suitable for governments with strict data regulations.

Governments migrating to Cloud will continue to require on-premises infrastructure. AUKUS partners will need access to a Cloud that enables organisations to navigate the transition while maintaining connections to legacy systems, ensuring a smooth evolution. Consequently, there will be a need to access a Cloud that excels in support for hybrid cloud scenarios, offers robust hybrid Cloud computing services capabilities, and is a bridge between on-premises environments and the Cloud.

AUKUS partners will need to communicate using a common productivity ecosystem, one already in wide use throughout governments. A platform that would improve productivity across the three nations through unified communication and collaboration providing features like persistent workplace chat, video meetings, file storage, and application integration. Ideally the migration and integration of these productivity services should be simple, and should be supported by robust AI features that combine the power of large language models (LLMs) with an organisation’s data.

AUKUS partners will need access to a Cloud that provides an array of best-in-class AI and machine learning services. A Cloud that enables AUKUS nations to partner with Industry – big or small – and is the go-to platform for a wide range of developers seeking to harness advanced analytics. A Cloud that will allow developers access to a platform to collaborate with their respective AUKUS partner, where they can securely and flexibly build, manage and deploy applications, drawing deeper insight from analytics, solve complex challenges, and share their outcomes across the AUKUS enterprise.

Azure Government Cloud’s comprehensive offerings, compliant security, and innovative AI services make it the right choice for an AUKUS Cloud solution. Azure’s migration process, as demonstrated by the successful transition of the US Department of Defense, supports hybrid cloud deployments, allowing for a smooth transition and integration with existing infrastructure. Azure also integrates seamlessly with Microsoft 365, enhancing collaboration and efficiency across the alliance. Furthermore, Azure offers an array of pioneering AI and machine learning services, such as Azure Machine Learning and Azure Cognitive Services, providing cutting-edge technology to support the alliance’s mission. A notable example is the integration of Microsoft’s AI assistant, Copilot, which can assist developers in writing code, thereby accelerating digital transformation efforts such as in the US military’s various software factories.²³ Critically, Azure Government Cloud is renowned for its compliance and security features, meeting strict US Government regulations.²⁴ These features ensure data protection and are suitable for governments with strict data regulations. Microsoft’s commitment to data sovereignty ensures that each member of the AUKUS alliance would maintain control over their data.  

Conclusion

To reach its potential, AUKUS needs access to Cloud. It is the gateway to exploiting innovative technologies and AI capability that will accelerate and transform the AUKUS partnership.²⁵ Cloud will provide the foundation on which AUKUS and industry partners will collectively build and deliver future capability. Working closely with industry, Cloud will enable the advanced applications and cost-effective services AUKUS needs at the speed of mission. By embracing a dedicated, secure Cloud solution AUKUS can keep pace with, and succeed against, any adversary through the delivery of on-demand AI services and applications that are easily accessible and rapidly scalable. Selecting the right Cloud will be critical to achieve these outcomes. Microsoft’s Azure Government Cloud is the right choice for AUKUS.

¹ van der Vlist, F., Helmond, A., & Ferrari, F. (2024). Big AI: Cloud infrastructure dependence and the industrialisation of artificial intelligence. Big Data & Society, 11(1). https://doi.org/10.1177/20539517241232630https://journals.sagepub.com/doi/10.1177/20539517241232630

² https://azure.microsoft.com/en-us/resources/cloud-computing-dictionary/what-is-artificial-intelligence/ Artificial intelligence (AI) is the capability of a computer system to mimic human-like cognitive functions such as learning and problem-solving.

³ https://assets.publishing.service.gov.uk/media/62a7543ee90e070396c9f7d2/Defence_Artificial_Intelligence_Strategy.pdf

⁴ https://www.bloomberg.com/features/2024-ai-warfare-project-maven/?leadSource=uverify%20wall

⁵ In computing, a workload is typically any program or application that runs on a computer. Workload can also refer to the amount of work — or load — that software imposes on the underlying computing resources. AI workloads are fundamentally different from the traditional data centre tasks. They rely on extremely high-performance computing nodes and are normally distributed across multiple CPUs and GPUs that need to communicate with each other in real-time.

⁶ https://azure.microsoft.com/en-us/resources/cloud-computing-dictionary/what-is-the-cloud/ Cloud computing is the delivery of computing services – including servers, storage, databases, networking, software, analytics, and intelligence to offer faster innovation, flexible resources, and economies of scale. The Cloud is not a physical entity, but instead is a vast network of remote servers around the globe which operate as a single ecosystem. These servers are designed by hyperscalers to either store and manage data, run applications, or deliver content or a service such as streaming videos, web mail, or office productivity software.

⁷ https://www.defense.gov/News/Releases/Release/Article/3239378/department-of-defense-announces-joint-warfighting-cloud-capability-procurement/

⁸ For more detail on the proposed AUKUS Cloud, see https://devblogs.microsoft.com/azuregov/accelerating-aukus/

⁹ https://www.c4isrnet.com/artificial-intelligence/2024/02/02/intel-community-tools-can-speed-ai-adoption/?utm_source=sailthru&utm_medium=email&utm_campaign=c4-overmatch

¹⁰ https://assets.publishing.service.gov.uk/media/63ce5a1b8fa8f53fe70e0857/Cloud_Strategic_Roadmap_for_Defence.pdf

¹¹ https://leidos.widen.net/s/jjxhjc2n9p/v5-future-intelligence-analyst-mar24

¹² https://aiindex.stanford.edu/wp-content/uploads/2024/04/HAI_2024_AI-Index-Report.pdf

¹³ https://www.aph.gov.au/Parliamentary_Business/Hansard/Hansard_Display?bid=committees/commsen/27911/&sid=0000

¹⁴ https://www.forbes.com/sites/williamhartung/2024/04/19/getting-past-the-hype-on-emerging-military-technologies-is-a-life-and-death-issue/?sh=5a4d3ed54717

¹⁵ https://www.defense.gov/News/Releases/Release/Article/3489803/dod-announces-establishment-of-generative-ai-task-force/ Led by the Chief Digital and Artificial Intelligence Office (CDAO), Task Force Lima will assess, synchronize, and employ generative AI capabilities across the DoD, ensuring the Department remains at the forefront of cutting-edge technologies while safeguarding national security.

¹⁶ https://www.c4isrnet.com/artificial-intelligence/2024/02/02/intel-community-tools-can-speed-ai-adoption/?utm_source=sailthru&utm_medium=email&utm_campaign=c4-overmatch

¹⁷ https://www.scsp.ai/wp-content/uploads/2024/01/human-machine-teaming-sr-jan-2024.pdf

¹⁸ https://www.gov.uk/government/news/chief-of-defence-intelligence-rusi-webinar-may-2023

¹⁹ https://www.gov.uk/government/news/chief-of-defence-intelligence-rusi-webinar-may-2023

²⁰ https://blogs.icrc.org/law-and-policy/2023/10/24/algorithms-of-war-use-of-artificial-intelligence-decision-making-armed-conflict/

²¹ https://assets.publishing.service.gov.uk/media/63ce5a1b8fa8f53fe70e0857/Cloud_Strategic_Roadmap_for_Defence.pdf

²² https://assets.publishing.service.gov.uk/media/63ce5a1b8fa8f53fe70e0857/Cloud_Strategic_Roadmap_for_Defence.pdf

²³ https://learn.microsoft.com/en-us/microsoft-copilot-studio/requirements-licensing-gcc

²⁴ For example FedRAMP High is the highest impact level in the U.S. Federal Risk and Authorization Management Program (FedRAMP). DoD SRG Level 6, also known as Department of Defense Impact Level 6 (IL6), is a security standard used by the U.S. Department of Defense (DoD) to assess the security posture of a cloud service offering (CSO).IL6 is reserved for the storage and processing of information classified up to the SECRET level.

²⁵ https://assets.publishing.service.gov.uk/media/63ce5a1b8fa8f53fe70e0857/Cloud_Strategic_Roadmap_for_Defence.pdf