Skip to main content

Operational AI Integration and Governance in the Australian Army

Journal Edition
DOI
10.61451/1235807

Overcoming Challenges for Strategic Advantage

 

Introduction

Contemporary technological advancements in the field of artificial intelligence (AI) bring to the forefront of global discussion the possibility of increasingly integrating adaptive AI technologies into battlefield operations. This discussion frequently takes the position that AI and autonomous technologies may revolutionise warfare and deliver asymmetric advantage to nations that utilise them effectively.[1] The often-proposed benefits of military AI and autonomy are numerous, including decision advantage, the safer and more efficient utilisation of personnel, the ability to generate mass and scalable effects, and enhanced force projection.[2] However, the greatest advantage of AI and autonomy for states and their militaries may be to unlock capabilities that can continually adapt to evolving operational and tactical problems, using innovation to seize the initiative and disrupt adversaries with transient technological advantages.[3]

There are unique complexities in the land domain that mitigate against the introduction of novel technological warfighting systems. This is because, unlike in the domains of maritime, air, space and cyberspace, non-AI-enabled software struggles to scale to the unbounded number of possible terrain-based and human variables in the land domain. AI and AI-enabled autonomous systems have particular utility in land operations because they can ‘learn’ how to navigate uncertainty, a characteristic which adversaries will struggle to overcome.[4] AI technology therefore presents significant potential for nations that are able to create and capitalise upon technological progress to deploy effective land-domain AI-enabled systems at scale. The possibility for these capabilities to contribute to Australia’s strategic priorities prompts the question: how can the Australian Army effectively operationalise AI in the land domain in order to maximise the strategic advantage it will provide in a future conflict?

This article explains the unique benefits and challenges to the Australian Defence Force (ADF) in its efforts to develop and use operational AI in the land domain. It argues that these unique factors are significant enough to warrant domain-specific governance considerations to ensure consistency with the broader Australian defence landscape. However, for the ADF to gain and maintain asymmetric advantage in the land domain using AI and AI-enabled autonomy, the technology needs to be reliable.[5] Armies, including the Australian Army (Army), risk operating at reduced effectiveness if the governance processes that influence land domain AI’s development and use are ill-suited to the operational context. After all, a commander will not use an AI-enabled system unless confident that its use will support achievement of the mission.

This article defines AI as ‘a collection of techniques and technologies that demonstrate behaviour and automate functions that are typically associated with, or exceed the capacity of, human intelligence’.[6] AI technologies, and their integration with human command and control, will be the primary driver of robotics and autonomous systems (RAS), enabling systems to reliably perform tasks or functions without (or with limited) direct human control, while continually adapting in highly dynamic environments.[7]

This article seeks to inform future efforts to apply and use AI technologies in the land domain, as well as related policy and governance frameworks. While significant progress has already been made in the research, development and demonstration of AI technologies, attention must now shift to the challenge of integrating AI within command and control architectures and developing governance mechanisms to enable its use in land domain operations.

Endnotes

[1] Michael C Horowitz, ‘Artificial Intelligence, International Competition, and the Balance of Power’, Texas National Security Review 1, no. 3 (2018): 38.

[2] Australian Army, Robotics and Autonomous Systems Strategy v2.0 (Canberra: Army HQ, 2022), pp. 7–18.

[3] Ash Rossiter and Peter Layton, Warfare in the Robotics Age (Boulder: Lynne Rienner Publishers, 2024), pp. 146–149, 162–163.

[4] Ibid., p. 33.

[5] Austin Wyatt, Joanne Nicholson, Marigold Black and Andrew Dowse, Understanding How to Scale and Accelerate the Adoption of Robotic and Autonomous Systems into Deployable Capability, Australian Army Occasional Paper No. 20 (Canberra: Australian Army Research Centre, 2024).

[6] Australian Army, Robotics and Autonomous Systems Strategy, p. 4.

[7] Ibid., p. 5.