From AI adoption to AI leadership

At CEDA’s State of the Nation 2026, artificial intelligence was not treated as a narrow technology issue. It appeared in discussions about productivity, manufacturing, research, energy, skills, sovereignty and trust.

The next phase of AI policy should be less about whether Australia adopts AI, and more about whether Australia builds the capability to benefit from it.

AI as part of the 2050 prosperity agenda

Prime Minister Anthony Albanese placed AI near the centre of Australia’s long-term economic challenge. In his opening address, he identified “the transformative impact of Artificial Intelligence” as one of the global shifts that will shape Australia’s future prosperity, growth and productivity.

Importantly, he did not frame AI only as a productivity tool. He argued that Australia should bring its “national values of fairness and opportunity to AI”, so that it grows the economy and strengthens sovereignty without fragmenting society or damaging the environment.

That is a useful framing because it sets a higher bar than adoption alone. The Prime Minister called for AI deployment that supports economic resilience, institutional trust and public benefit.

From a taker to a maker

The clearest statement of an AI industry agenda came in the Research and Development for a Future Made in Australia session.

Minister Tim Ayres argued that Australia should approach AI “not as just a taker of other people’s technology, but as a maker of technology.” He connected this to a larger ambition: building more of the AI ecosystem onshore, including critical minerals, data centres, specialist chip components, training, norms and industrial applications.

That is a significant point. Australia does not need to match the United States or China in frontier model development. But it does need to decide where building domestic capability matters. These include AI applications in advanced manufacturing, resources, health, science, public services, logistics, energy and professional services.

Ayres also linked AI to the National Science and Resilience Council, the National Reconstruction Fund, the investor front door and Australia’s research system. The underlying point was coordination. AI capability will not emerge from isolated pilots or scattered grants. It will require a pipeline from research to commercialisation, and from commercialisation to scaled adoption.

That is one of the key questions for the AI Leadership Summit 2026: where does Australia need sovereign depth, where should it partner internationally, and where should it focus on fast, trusted adoption?

Skills Will Decide Whether AI Diffuses

The R&D session at State of the Nation also made clear that AI leadership is not only about frontier firms. Jenny Dodd’s contribution on TAFE and workforce capability pointed to a quieter but equally important issue: diffusion.

Productivity gains from AI will depend on whether businesses and workers can actually use the technology well. That requires more than a small group of technical experts. It requires managers who understand workflow redesign, workers who can use AI-enabled tools, educators who can adapt training quickly, and institutions that support lifelong learning.

Dodd’s discussion of TAFE Centres of Excellence, micro-credentials and place-based industry partnerships is relevant here. AI adoption will not occur evenly across the economy. Large firms may move quickly. Smaller firms may lack capability. Regional industries may need practical, applied support. Workers already in the labour market will need retraining, not just new entrants.

That should be central to the AI Leadership Summit agenda. If AI is treated only as a boardroom or research-lab issue, Australia will miss the main productivity challenge: broad adoption across firms, sectors and regions.

AI has an infrastructure problem

State of the Nation also surfaced an issue that is sometimes missing from AI discussions: the physical economy behind the digital one.

In the energy discussion, Liz Westcott pointed to AI and data centres as part of a broader increase in energy demand. The Prime Minister also referred to Microsoft’s new data centres in New South Wales in the context of large energy users investing in renewables.

This is a useful corrective. AI is not weightless. It requires data centres, energy, transmission, cooling, chips, minerals, capital and skilled labour. If Australia wants to expand AI capability, it must also think seriously about the infrastructure that makes AI possible.

That links AI policy to clean energy, grid reliability, planning approvals, critical minerals, land use and investment certainty. The AI economy will not be built only in software. It will be built through energy systems, industrial capability and infrastructure planning.

Towards the AI Leadership Summit 2026

The AI discussion at State of the Nation did not produce a single AI blueprint. But it did point to the right questions.

Can Australia move from AI adoption to AI capability? Can it build firms that scale, not just firms that experiment? Can it equip workers and managers to use AI well? Can it secure the energy, data and infrastructure required? Can it govern AI in a way that builds trust rather than eroding it? And can it connect AI to national priorities rather than treating it as a technology trend floating above them?

That is the opportunity for the AI Leadership Summit 2026.

The Summit should not only ask what AI can do. It should ask what Australia needs to do so AI strengthens the economy, supports the social compact and builds national capability.

State of the Nation made one thing clear: AI is now part of the bigger prosperity question. The task ahead is to make sure Australia does not merely consume the next wave of technology, but shapes it in ways that serve the country’s long-term interests.