The release of the United States’ AI Action Plan marks a shift in tempo. It’s now abundantly clear Washington is treating artificial intelligence as a pillar of industrial policy, national defence and technological competition. The plan consolidates several strands of US AI activity into a single strategy that aims to accelerate domestic capability while shaping international norms and supply chains.
For Australia, the plan presents both opportunity and pressure. It creates new avenues for collaboration across AUKUS Pillar Two, Quad tech initiatives and global standard-setting efforts. But it also potentially increases expectations. As the US sharpens its doctrine, partners may well will be expected to demonstrate strategic alignment not just in principle but in policy, infrastructure and enforcement.
The plan promotes open-source model development, investment in computing capacity and chip manufacturing and the expansion of a domestic AI workforce. It also outlines new guardrails, including export controls on high-risk models and restrictions on technologies linked to authoritarian regimes. Particular attention is given to countering the spread of artificial intelligence systems developed under the influence of the Chinese Communist Party. This includes efforts to identify and block models that reinforce state propaganda, enable digital repression or support military planning.
The challenge for Australia is how to keep pace. Positively, the Albanese government has already signalled an interest in sovereign compute, risk-based regulation and responsible AI development. Coordinated infrastructure, strong governance and clear institutional leadership will also be required to ensure those priorities don’t fragment. This is why the government’s development of a National AI Capability Plan is both so welcomed, but also necessary. Partners are watching closely. Interoperability, verification and trust are fast becoming the measures of alignment.
Export controls are one area where expectations are changing. The US plan calls for tighter scrutiny of foundational models and training data, especially where systems may be repurposed for military or coercive use. The question for Australia to consider is whether our current screening mechanisms are sufficient. There is a growing need for tools that can track the provenance of large models, monitor for misuse, and manage investment risk in high-exposure areas. This includes dual-use compute infrastructure, datasets with potential for surveillance applications, and technology transfers through intermediary markets.
Deregulation will also draw attention. The US plan includes provisions for innovation zones and streamlined approval pathways, which may encourage industry groups in Australia to push for a lighter regulatory touch. We should expect the government’s National AI Capability Plan will be clear about Australia’s risk posture. The challenge is how to do so with sufficient nuance that we have an appropriate sectoral based approach rather than a one-size-fits all model. This recognises a basic fact: a coherent AI regulatory framework needs to allow innovation to thrive while also protecting system integrity, public safety and international confidence.
The US commitment to AI infrastructure should be expected to have regional effects. Large-scale investments in data centres, energy capacity and chip fabrication are already influencing supply chains across the Indo-Pacific. Australia may face increased competition for skilled labour, capital and materials as other countries move quickly to secure strategic advantage. Without early action, Australia risks losing ground in areas such as sovereign compute, trusted cloud services and the deployment of secure AI systems in critical sectors.
This strategic realignment is not limited to the US. Countries such as Japan and Canada are also expanding their AI posture, with a focus on democratic governance, supply chain resilience and regional cooperation. Japan’s role in shaping Indo-Pacific digital norms continues to grow, while Canada’s investments in public infrastructure for AI provide a model for accessible and trusted systems. Deepening engagement with these partners is a force-multiplier. Aligning with a broader coalition of like-minded states can help reduce dependency, diversify capability and reinforce shared principles in a contested region—positioning not just Australia but also these partners well in the context of harshening technology competition between the US and China.
Workforce development remains a priority. The US plan expands support for technical training across AI-adjacent fields, not just software engineering. This includes trades, infrastructure and operational roles that will be essential to deploying and maintaining AI systems at scale. A similar approach would benefit Australia. National capability will depend on a wide talent base, not just elite technical expertise. Opportunities also exist to participate in US-led education exchanges, research initiatives and mobility programs—provided Australia is ready to plug in.
The US has made clear that artificial intelligence is now a strategic domain. Capability will be measured by what countries can design, deploy and secure in practice. The government’s National AI Capability Plan, which is under development, is an opportunity to respond with equal clarity. That means aligning institutions, clarifying governance and investing in the infrastructure needed to deliver trusted systems at scale.