Australia Should Have a Stake in the AI Economy

Australia is considering how to build AI infrastructure that works for Australia. It is a necessary conversation, and it is much bigger than data centres.

It is about what kind of AI economy we are designing.

At Wonder & Wander, we have made a submission to the Australian Government consultation on AI infrastructure with one simple position:

Australia should not just host the AI economy. We should have a stake in it.

If Australian land, energy, water, data, talent and communities help enable the next generation of AI, Australians should share meaningfully in the capability and value created.

We are pro-AI. We are pro-innovation. And we are interested in making sure Australians are not simply consumers of a future built around us.

The infrastructure question is also a sovereignty question

Onshore infrastructure matters. It can improve resilience, support secure data handling and create room for Australian research and industry.

But physical location alone does not create sovereignty.

The bigger question is whether infrastructure in Australia translates into capability that Australians can access, understand, influence and benefit from. For strategically significant infrastructure, we believe the national-interest test needs to consider more than megawatt capacity. It needs to consider assured access, ownership and control, jurisdictional exposure, domestic capability and the economic value that stays here.

Our submission frames this through four connected ideas.

01. Infrastructure sovereignty

Australia should retain assured access to critical compute and infrastructure, including during disruption or scarcity. That does not necessarily mean government needs to own every facility. It does mean critical services, public-interest research, Australian startups and nationally significant AI development cannot be entirely dependent on decisions made elsewhere.

A useful question is: if a provider changed commercial priorities, access conditions or jurisdictional arrangements tomorrow, would Australia still have access to the capability it needs?

02. Data sovereignty

Data sovereignty is about clear governance over where Australian data lives, who can access it, which jurisdiction applies and how it can be used.

For strategically significant facilities and AI training activity, Australians should be able to understand where data is stored and processed, the conditions under which it might move, who can access it, and whether it can be used for model training or secondary purposes.

This is not an argument for digital isolation. Australia can participate deeply in global technology ecosystems while retaining clear national protections and the ability to act in the public interest when required.

Data sovereignty is national agency.

03. Economic sovereignty

Australian infrastructure investment should create Australian research, IP, businesses, jobs, startups and commercial value.

That is why we have proposed an Australian Capability Dividend: a measurable expectation that major AI infrastructure and frontier-AI investment leaves Australia more capable than it found it.

It could include research partnerships, local engineering and AI teams, apprenticeships, compute access for startups and researchers, Australian-owned IP, local procurement and commercialisation pathways. The measure is not attendance at a training session. It is what people and organisations can do afterwards: build, redesign, create, earn, lead and own.

04. Human sovereignty

AI should expand human capability and agency, not quietly remove it.

As AI becomes embedded in work, finance, healthcare, education, government and everyday life, Australians should retain meaningful rights to know, understand, challenge, participate and exercise human judgement.

We call this digital citizenship.

Digital citizenship gives us a practical way to ask whether people can know when consequential AI is being used, understand how it affects them, challenge important decisions, exercise appropriate control over their information, avoid unnecessary lock-in, participate in new opportunity and remain meaningfully human in systems that shape their lives.

Digital citizenship is individual agency.

From host community to capability community

The consultation also raises an important question about communities hosting major AI infrastructure.

“No net costs” should be the floor. Communities should not be asked to carry additional energy, water, infrastructure or social costs for a project that creates value elsewhere. But the ambition can be stronger.

Our companion paper proposes a Community Capability Dividend: a measurable commitment that major AI infrastructure projects leave behind stronger local skills, businesses, digital capability, innovation capacity, participation and agency.

This is not a community grants program. It is a design principle for infrastructure.

The question changes from how do we compensate a community for impact? to what durable capability will exist here because this investment happened?

The model explores six possible categories of dividend:

• Economic dividend: local procurement, supplier development, regional business growth and long-term jobs.
• Workforce dividend: apprenticeships, technical pathways, reskilling and careers that continue beyond construction.
• Digital dividend: connectivity, safe AI access and practical AI capability for local organisations.
• Innovation dividend: SME pilots, startup pathways, research partnerships and access to expertise or compute.
• Community dividend: investments shaped by local priorities, including education, resilience, services and quality of place.
• Ownership and agency dividend: meaningful participation in decisions and, where appropriate, long-term economic value.

The approach starts with lived experience. Communities are not audience segments. They are living systems of households, businesses, workers, schools, councils, First Nations organisations, service providers and informal networks. Good engagement begins by listening, then co-designing commitments with the people who will live with the outcomes.

An AI Native Innovation Fund

We have also proposed exploring an AI Native Innovation Fund to help Australian founders, researchers, SMEs and communities turn access to AI infrastructure into Australian products, IP and companies.

Infrastructure becomes national capability only when people can use it to make, test, learn, commercialise and lead. A future-ready agenda should create clear pathways from access to application, from learning to ownership, and from infrastructure to enduring Australian value.

The opportunity

Australia does not need to choose between attracting global AI investment and protecting national capability. We can do both deliberately.

The goal is not to slow progress down. It is to design the conditions under which progress is genuinely shared.

Australia should help build the future. Australians should benefit from it. And Australians should have a stake in it.

The two papers below set out our complete submission and the companion Community Capability Dividend framework in more detail.