In the 1980s and 1990s, Labor governments under Bob Hawke and Paul Keating confronted an uncomfortable reality: Australia could not drift into prosperity. The old economic settings — protection, inertia and closed markets — were failing Australia. So Labor floated the dollar, rewired the economy, opened Australia to the world and reimagined us as a competitive, outward looking nation.

These reforms were fiercely contested at the time, but they worked. Labor did not resist disruption — it shaped it in the national interest, all the while enhancing access, opportunity and Australia’s fair go.

Australia now faces a similar moment. A new general purpose technology has arrived, and once again the question is not whether we adopt it, but whether we shape it — or simply become a customer of it.

Every country tells itself the same comforting story about new technology. We’ll adopt quickly. We’ll be world leading users. Prosperity will follow. Australia told itself that story in the 2010s. We became the most Uber dense market in the world. We embraced Google and Facebook faster than almost any comparable economy. We were exceptional adopters.

And we got cleaned out.

Today, around 71 cents of every digital advertising dollar spent in Australia flows to Google or Meta, both headquartered in California. Uber controls roughly 80 per cent of our rideshare market — while the profits accrues offshore. Australians got convenience. Foreign shareholders got the value. This is the Uberisation of the Australian economy, and it should echo through every major procurement decision being made right now.

Because the next wave is here — and it is much larger.

Artificial intelligence is not just another technology. It is the layer on which every other technology will be built. Over the next decade, AI is expected to add between US$4 trillion and US$15 trillion to global economic output — a prize comparable to the internet itself. Australia is standing at a sliding doors moment, where two futures diverge based on the choices we make in the next few years.

The difference between those futures is ownership.

Consider a simple thought experiment. An Australian worker is given an AI tool that can do much of their job. They deploy it, productivity rises, the business benefits, and the worker is freed to pursue higher value tasks. Everyone wins.

Now imagine a second scenario. Instead of simply adopting the tool, the firm builds the AI itself. The work still gets done, but the gains from higher productivity are no longer automatic. Workers may need to retrain or redeploy, and outcomes depend on how deliberately the technology is integrated. Progress is still possible — but it must be shaped.
Now imagine a third scenario. The AI is owned by a foreign monopolist, sold to Australian employers at a price just below the cost of wages. The worker is displaced. The firm barely improves its margins. And the economic value flows offshore.

Same technology. Three radically different outcomes. Ownership is not a side issue — it is fundamental to national welfare. And this is not theory. It is already happening.

Consider Harrison.ai, one of Australia’s most compelling AI success stories. Harrison.ai builds clinical grade artificial intelligence that helps doctors detect disease earlier and more accurately. Today, its technology is deployed across more than 1,000 healthcare facilities in around 40 countries. It is live across dozens of NHS Trusts in the United Kingdom and has raised over $300 million including investment from the Australian Government’s National Reconstruction Fund. It is a globally credible Australian AI firm.

But here is the critical point: Harrison.ai easily could have gone the other way.

Its breakthrough was not inevitable. It was not guaranteed by superior Australian talent alone. It happened because an Australian radiology network took a chance on a local company early — when the technology was promising but not yet proven at scale.
That single decision mattered. It created early revenue, clinical validation, training data and global credibility. It generated momentum. Without it, Harrison.ai might never have reached escape velocity. The same engineers might have been quietly acquired by a foreign firm — or simply left Australia altogether.

That is how close these calls can be.

Multiply that moment by hundreds — across health, logistics, energy, education, defence and public administration — and the direction of Australia’s AI future becomes clear. Either we build an AI industry of our own, or we entrench permanent dependence on foreign platforms.

Building world-leading foundation models requires hundreds of billions of dollars in compute. But applied AI is a different game. It is where deep industry expertise meets trusted local data. And in that contest, Australia competes at a world class level.

We already have more than 1,500 AI-focused companies. We have globally respected research institutions. We have firms like Lorikeet in customer intelligence, Heidi Health in clinical documentation, and Relevance AI in workflow automation competing internationally. They are highly capable. What they lack is access to anchor customers willing to put Australian firms on the shortlist.

This is not an argument for protectionism. No-one should buy inferior technology out of nationalism. But when the choice is close — and often it is — Australia should lean Australian. Because the compounding effect of hundreds of such decisions will determine whether we build sovereign capability or spend the next generation renting intelligence from abroad.

Hawke and Keating understood something essential: markets alone do not decide a nation’s destiny. Choices do. The doors are open now — but they will not stay open forever. Countries that build AI will capture its value. Countries that merely use it will become dependent on those who do.

Australia has earned the right to be a builder again.

We just have to choose it.