AI, Cybersecurity, and Defense: The West’s Great Gamble and Australia’s Critical Test

The global arms race has pivoted from tanks and aircraft to algorithms and neural networks. As defense systems across the West are rapidly infused with artificial intelligence, the rewards—faster threat detection, autonomous battlefield responses, predictive analytics—are matched only by the magnitude of risk. For Australia and its allies, the imperative is no longer simply to innovate, but to secure and control. Without that, the very systems designed to protect national sovereignty could spiral into liabilities.

According to Grand View Research, the military cybersecurity market—valued at $14.05 billion in 2024—is projected to surge past $52 billion by 2034. At the center of this transformation is AI: reshaping military doctrine, accelerating cyber-operations, and unlocking a new industrial boom for defense contractors.

Yet as the tempo of deployment increases, so does the vulnerability of critical systems to manipulation, failure, and misuse.

“AI-enabled systems are highly vulnerable to cyber attacks in ways that traditional military platforms are not,” warns Alice Saltini, Non-Resident Expert on AI at the James Martin Center for Nonproliferation Studies. “These systems create new entry points for hackers to access and manipulate sensitive military data or disrupt operations.”

Alice Saltini on the implications of AI in nuclear command, control & communications. YouTube.

Her comments, published in a February 2025 report for the European Leadership Network (ELN), underscore a chilling truth: AI in defense is a double-edged sword. And it’s not being sharpened evenly.

The UK Sounding the Alarm

The United Kingdom and Australia are not alone in grappling with the implications. In the UK, former First Sea Lord and former Security Minister Lord West of Spithead has gone public with his deep concerns over AI’s role in nuclear command, control, and communication systems (N3).

“We must ensure that human political control of our nuclear weapons is maintained at all times,” Lord West said bluntly in an interview with The Canary.

“Let’s be very wary if we do anything in this arena of AI, because the results could be so catastrophic. It would be very nice to have some more clarity about this, and some more reassurance about the work that’s actually going on.”

West’s appeal is a direct challenge to policymakers across the Western alliance: commit to transparent oversight, or risk a future where human error is replaced by machine misjudgment—with irreparable consequences.

Despite the UK’s earlier creation of the AI Council as an independent advisory group, its dissolution and replacement by the less governance-focused AI Safety Institute has left oversight fragmented. This signals a wider issue: while AI adoption in defense is accelerating, formal regulatory structures are falling behind.

*Royal Australian Navy, showing Deputy Prime Minister and Minister for Defence of Australia Richard Marles.*

Australia’s Position: Opportunity and Vulnerability

Australia, through the AUKUS pact and growing engagement with the Quad, is well-positioned to become a leading node in allied AI defense infrastructure. The Australian Signals Directorate (ASD) and the Defence Science and Technology Group (DSTG) are actively investing in machine learning for threat detection and cyber resilience.

But crucial gaps remain. Without a national AI risk management body—or a security architecture equivalent to the Five Eyes-level oversight of sensitive technologies—Australia risks importing not just capabilities, but systemic vulnerabilities.

Meanwhile, Australian startups and defense firms are ramping up partnerships, aided by funding via the Defence Innovation Hub and collaborative agreements with U.S. and Israeli cyber-defense innovators. This momentum must be matched by a whole-of-nation strategy for AI risk controls.

Strategic Growth—And Strategic Risk

Defense titans led by Raytheon, Northrop Grumman, and Lockheed Martin are already deploying AI to automate battlefield surveillance, analyze enemy behavior in real time, and support logistics decision-making. These systems, often run through black-box algorithms, operate at speeds and scales incomprehensible to human controllers.

But these benefits come with a price: adversarial actors such as China and Russia are weaponizing AI themselves. From deepfake propaganda and automated phishing to AI-managed drone swarms, cyber warfare is now being conducted at both psychological and operational levels.

“AI has power-seeking tendencies,” Saltini cautions. “It can optimize toward goals we don’t fully understand or control, particularly in black-box neural architectures.”

In this context, it is no longer enough to ask if AI works—we must ask how it can fail, who it can mislead, and how it can be corrupted.

Global Implications and the Path Forward

The global implications of militarized AI are escalating rapidly. Adversaries such as China and Russia are intensifying their cyber warfare operations—deploying sophisticated espionage tactics to infiltrate critical infrastructure and defense networks.

The U.S. Department of Defense has flagged China as a growing cyber threat, citing its expansive efforts to steal strategic data and disrupt Western military readiness. Russia, for its part, has already demonstrated its digital warfare capabilities, from the 2016 U.S. election interference to cyberattacks designed to destabilize geopolitical rivals.

Both nations are almost certainly probing Western AI-integrated systems for exploitable gaps, especially in moments of geopolitical tension. The pre-conflict window—where digital signals and defense posturing play a decisive role—is now a prime arena for AI-led incursions.

To counter this, Western allies must implement hardened systems, including quantum-resistant encryption, zero-trust cybersecurity architectures, and human-machine teaming that enhances AI responsiveness without sacrificing judgment. Collaboration across borders is essential—shared intelligence, standardized protocols, and coordinated responses must become the norm, not the exception.

As Alice Saltini stresses, AI-enabled defense systems come with "unique vulnerabilities" that demand urgent, structured attention—not after a breach, but well before one.

Australia’s participation in AUKUS Pillar II offers a pivotal platform to scale up its cyber capabilities. Some defense experts argue it could become the ideal mechanism for bolstering workforce development and cyber-operational capacity in line with escalating threats.

“Cyber Warfare Specialists and Cyber Warfare Officers are core to the conduct of cyberspace operations that detect and defeat cyberspace attacks against Defence networks and systems,” said Lieutenant General Susan Coyle, Chief of Joint Capabilities.

“The new Cyber Warfare Common Remuneration Framework will enable cyber personnel to be employed within a skills-based pay structure—better addressing retention and the deep technical skills needed in this domain.”

This evolving cyber workforce strategy—combined with AI threat modeling and resilience planning—should serve as a call to action. Canberra must now lead on the global stage, pushing for the development of an AI arms control regime—a 21st-century analogue to nuclear non-proliferation treaties.

Anchored by Five Eyes intelligence and framed around transparency, interoperability, and shared accountability, such a framework would be a crucial step in managing the existential risks AI now poses.