The AI scorecard Australia did not know it needed

Most conversations about artificial intelligence in Australia start from a position of anxiety. Are we falling behind? Are we too dependent on foreign technology? Do we even have a seat at the table?

New research suggests the answer is more encouraging than the prevailing narrative allows. Australia’s first comprehensive, independent, evidence-based assessment of the country’s AI capabilities finds that Australia holds high or very high agency across the vast majority of the 103 AI capabilities measured.

Low agency exists in only two. For a country that often talks itself down on technology, that is a result worth paying attention to.

What the research actually measured

The report, titled Expanding AI Sovereignty to AI Agency, was released by the Tech Policy Design Institute (TPDi). It was developed in consultation with more than 250 experts and maps Australia’s AI capabilities against the Australian Government’s 2025 National AI Plan.

AI sovereignty and AI agency sound similar. They are not. Sovereignty implies self-sufficiency, the idea that a country should own and control every part of its AI stack. Agency is a more useful concept. It measures the power of a country to shape its own AI future, even within a globally interdependent system.

TPDi Co-Founder and Executive Director Johanna Weaver make the point directly. “The debate about AI sovereignty has become trapped in a choice between complete self-sufficiency and complete dependence. This is a false binary. The reality is that the AI supply chain is a complex global web. Every country relies on others.”

The difference between the two concepts is where the real opportunity sits. Instead of asking whether Australia can go it alone, the more useful question is where Australia AI agency is actually built. The answer lies in deliberate, strategic positioning within a global system rather than isolation from it.

Where Australia actually leads

The assessment finds Australia holds very high agency in eight capabilities and high agency in another 58. The areas of genuine strength include geospatial data, international influence, data centre infrastructure and public cloud. These are not peripheral capabilities. They sit at the foundation of how AI systems are built and deployed at scale.

Zoe Jay Hawkins, TPDi Co-Founder and lead author, says the findings challenge a narrative that has undersold Australia’s position.

“The data shows Australia is in a stronger position than we give ourselves credit for. We have firm foundations and significant potential to harness.”

The report also identifies six high-agency capabilities that the Government’s National AI Plan has not yet fully committed to. These include investment in public sector and public interest compute, inclusive AI adoption, culturally and nationally inclusive models, and regulatory oversight. The research flags these as priority areas for the next phase of policy development.

The strategic opportunities hiding in plain sight

Beyond the capability mapping, the report identifies three areas where Australia’s existing strengths could be leveraged more aggressively.

Australia’s potential as a regional data centre hub could help fund the country’s clean energy transition. Australia’s critical minerals reserves could secure supply of the advanced AI chips required to power the future economy.

And Australia’s data assets, which the report describes as among the richest in the world, could be deployed to deliver better emergency management and health outcomes.

Hawkins is direct about what the research enables. “This research allows decision-makers to understand the full chessboard of AI capabilities and make decisions about which to prioritise in the national interest, based on evidence rather than spin.”

That applies to the private sector as much as government. For investors and asset managers thinking about AI exposure, understanding Australia AI agency within the global supply chain is increasingly relevant to portfolio construction and thematic positioning.

Not every gap is a problem

One of the more nuanced findings in the report is that low agency in certain capabilities is not necessarily a failing. Australia scores low on manufacturing AI chips and building frontier models.

However, the assessment finds these are areas where Australia has low competitive advantage anyway. Chasing them would be a misallocation of resources. Chasing them would be a misallocation of resources.

“The assessment allows us to understand our strengths and more deliberately leverage them to fill our weaknesses,” Hawkins says. “Australia has valuable cards in its hand. The opportunity now is to play them strategically.”

What this means for investors and advisers

AI is not just a technology story. It is an infrastructure story, an energy story, a data story and increasingly a geopolitical story. Australia sits at the intersection of several of these threads in ways that are not yet fully reflected in how the investment community thinks about domestic opportunity.

Understanding Australia AI agency is no longer just a policy conversation. It is a framework for identifying where genuine, durable investment opportunity sits in the decade ahead.

The TPDi report is worth reading in full. It is available at tpdi.org.au and provides the most rigorous, evidence-based picture of Australia’s AI position yet produced. For advisers and investors trying to cut through the noise on AI, it is a strong starting point.