Australia’s AI Data Centre Bet: Are We Flying Blind?
The Government wants Australia to become a regional AI infrastructure hub. But where is the national business case, where was the public consultation, and how do we measure whether it is actually in Australia's national interest?
This is a SUNA opinion and analysis piece. It draws on research undertaken for our Data Centres briefing paper, together with publicly available government, industry and economic data. It reflects our analysis of Australia's emerging data centre strategy and the questions we believe Australians should be asking about its long-term national benefit.
The Australian Government has already decided it wants Australia to become a regional AI and data centre hub.
That's a very big national ambition.
It could mean billions of dollars of international investment and the development of an entirely new Australian export industry.
It could also mean committing enormous quantities of Australian electricity, water, land and infrastructure to servicing the computing needs of international technology companies and customers for decades.
So before we go much further, there are some fairly obvious questions.
Where is the national business case?
Has there been proper, comprehensive consultation with the Australian people?
How will Australians know whether the strategy actually delivered what was promised?
And perhaps most fundamentally: what does the Government actually mean by the "national interest"?
The Government has identified some things it considers part of Australia's national interest — including national security, data sovereignty and benefits to the Australian economy, people and communities.
But it has not established a clear, transparent and measurable framework for determining whether a major data centre actually passes that national-interest test.
That's not a minor omission.
It's the test against which the entire strategy should be judged.
Without it, we really are flying blind.
Investment isn't a business case
Governments love investment announcements.
Billions of dollars committed. Major international companies choosing Australia. New projects, construction jobs and infrastructure.
The numbers sound impressive.
But an investment announcement is not a national business case.
A national business case asks something much more difficult:
What are Australians putting in, what are Australians getting out, what are the alternatives, and what happens if the assumptions are wrong?
We already have an early example of why that distinction matters.
In the March quarter of 2026, the Australian Bureau of Statistics reported that investment in data centre machinery and equipment was the largest contributor to growth from business investment.
That sounds like an extraordinary economic success story.
But there was another part to it.
The ABS also reported that the majority of those capital assets were imported. The increase in investment therefore came with an increase in imports, partly offsetting its contribution to GDP growth.
There, in a single set of national accounts, is the problem.
Investment occurring in Australia is not the same thing as value being created and retained in Australia.
That's not an argument against foreign investment.
It's an argument for measuring what Australia actually receives from it.
If we're going to commit substantial national resources to becoming a regional AI infrastructure hub, Australians should be able to see the national business case.
Not the developer's business case.
Australia's.
National interest can't mean "because we said so"
Government repeatedly tells us its AI and data centre policies will advance Australia's national interest.
Fine.
Show us the test.
The Government has identified some of the things it considers important: national security, data sovereignty, economic benefits and benefits for Australian people and communities.
But we still don't know how those things are weighed, what evidence is required, or at what point a project consuming significant Australian resources would be judged to return too little to Australia to proceed.
That matters because these projects involve trade-offs.
A project consuming hundreds of megawatts may produce significant investment and tax revenue, but relatively modest permanent employment. A foreign-owned facility may bring capital into Australia while much of its sophisticated equipment is imported and its profits ultimately flow elsewhere. Electricity committed to an export-oriented data centre cannot simultaneously support another industry without additional generation.
None of those things automatically makes a project contrary to Australia's interests.
But somebody needs to weigh them.
How should permanent employment, tax revenue, Australian participation, domestic capability, ownership, research and intellectual property be weighed against electricity, water, land and infrastructure requirements?
And at what point would government say:
This project consumes too much and returns too little to Australia.
Because a test that can never be failed isn't much of a test.
If "national interest" is going to justify decisions involving Australia's resources for decades, Australians should know what it means before those resources are committed — not after.
And where was the national conversation?
There has been limited consultation around artificial intelligence, regulation and individual developments. The Government also says its data centre expectations were informed by industry, utilities, states and territories, unions and special-interest groups.
But we have found no evidence of a dedicated national public consultation on the much bigger strategic proposition:
Should Australia become a major regional AI and data centre hub, and on what terms should Australian resources be committed to achieving it?
That's a different question.
Australians may be invited to comment on individual data centre developments through state planning and environmental processes.
Those processes matter.
But consulting a community about noise, traffic, water use or the environmental impact of a particular development is not the same as consulting Australians about the national strategy.
By the time we're debating those things, a much bigger decision has already been made:
Australia is going to pursue large-scale data centre investment and position itself as a regional AI infrastructure hub.
When did Australians have that conversation?
Where were we shown the likely cumulative demand for electricity, water, land and transmission infrastructure alongside the expected national return?
Where was the discussion about how much infrastructure Australia needs for itself and how much additional capacity we should build to service international customers?
Perhaps Australians would support the strategy overwhelmingly.
Perhaps, presented with a compelling national business case and appropriate safeguards, most people would regard it as an extraordinary opportunity.
But meaningful consultation should include the strategy itself, not simply the consequences of implementing it.
There is a profound difference between asking Australians how something should be built and asking whether it should be built at this scale, for this purpose and on these terms.
Where is the national picture?
This becomes even more important because data centres are generally considered project by project.
One development might require hundreds of megawatts. Another might require hundreds more. Each can move through its own planning, connection and approval processes.
But Australia's electricity system, water resources, transmission networks and communities experience the cumulative result.
As we discussed in our first article, 11 large data centre projects representing around 5.4GW of maximum demand were already progressing through the National Electricity Market transmission connection process in the March quarter of 2026.
Not all of that capacity will necessarily be built or consumed.
But that's precisely why Australia needs a national picture.
We need to know what is proposed, what is credible, what has been approved, what is under construction and what is actually operating. Just as importantly, we need to understand how much of the eventual capacity is intended to meet Australian needs and how much is being built to service international demand.
At present, Australians cannot readily answer those questions from one authoritative national source.
Yet government is already pursuing the hub strategy.
Again, the sequencing seems backwards.
Why doesn't this dashboard already exist?
Perhaps the strangest thing about SUNA proposing a National Data Centre Dashboard is that much of the information required to build one should already exist.
Large data centres don't simply appear.
They pass through planning systems. They seek enormous electricity connections. Water requirements are assessed. Environmental impacts are documented. Infrastructure needs are considered. Governments announce major investments and actively facilitate projects.
Different agencies therefore hold different pieces of the puzzle.
Yet Australians cannot open a single national source and see what all those projects add up to.
Instead, researchers, journalists, communities and interested Australians are left trying to reconstruct the national picture from planning portals, network reports, company announcements, investor presentations and government documents.
That is essentially what we had to do while preparing SUNA's Data Centres briefing paper.
Why?
If Australia is deliberately pursuing a strategy to become a regional AI and data centre hub, surely assembling the national picture should have been one of the first things government did.
Perhaps government already has a comprehensive internal picture.
If so, publish it.
If it doesn't, that's arguably more concerning.
Either way, Australians should not have to reconstruct a national infrastructure strategy project by project to understand what is being committed in their name.
Turn the lights on
This is why SUNA's Data Centres briefing paper proposes a National Data Centre Dashboard.
It doesn't require another glossy strategy or another industry roundtable.
It requires a public baseline and a way of measuring what actually happens.
Australians should be able to see how much data centre capacity is proposed, approved, contracted, under construction and operating, alongside the cumulative electricity, water, land and infrastructure commitment.
But the dashboard should not stop at resource consumption.
It should show what Australia receives in return: permanent employment, Australian procurement and ownership, tax contribution, research and skills development, access to computing capacity and other measurable national benefits.
And crucially, it should follow projects after they open.
If a project promises jobs, measure the jobs actually created. If Australian procurement is promised, measure what occurred. If new renewable generation or water-efficiency improvements form part of the case for approval, publish the outcome. If substantial public infrastructure is required, show Australians what it cost.
That is how we move from promises to evidence.
And it would finally allow Australians to compare what individual projects were supposed to deliver with what the industry is delivering collectively.
Transparency shouldn't frighten a successful strategy
There is nothing anti-data-centre about any of this.
Australia needs secure digital infrastructure. We need access to advanced computing. Australian businesses, researchers and government need to participate in the AI economy.
International investment may also create a valuable new export industry.
If becoming a regional AI infrastructure hub produces the extraordinary benefits its advocates predict, a National Data Centre Dashboard will help demonstrate that success.
Government should welcome that.
Industry should too.
Because the bigger Australia's data centre ambition becomes, the less acceptable it is to measure success through investment announcements and projected benefits alone.
We need to know what Australia actually receives and what resources we actually commit.
And we need a national-interest test against which those outcomes can be judged.
The Government has already made its ambition clear.
Now Australians deserve to see the case behind it.
Define the national interest. Publish the national picture. Measure the promises against the outcomes.
Because if Australia is going to make one of the largest infrastructure bets of the AI era, we should at least turn the lights on before we place the bet.