Around the world, the core nature of artificial intelligence is settling into three geographical blocs. In Silicon Valley and Washington DC, AI is considered a tool for productivity and economic growth, and from there, a means of economic leverage. In Brussels, the focus is on protective regulation to manage technological and social risks. Meanwhile, Beijing is seeing this as a practical technology and is building alternative technology and international coalitions, marshalled through the World AI Conference (WAIC), to aggregate state-led technological influence across the Global South.
Australia has an opportunity to engage in a different and much more imaginative conversation about AI.
Throughout history, societies have built new forms of intelligence infrastructure. Writing extended memory beyond the individual. Libraries expanded access to knowledge. Public schools democratised learning. Universities accelerated discovery. The internet connected knowledge globally. AI represents the next layer of intelligence infrastructure.
The question for Australia is not whether AI arrives. The question is, what will we allow AI to do for us, with us and to us? Instead of viewing AI as a commercial opportunity (or threat), a regulatory conundrum, or an instrument of global influence, what if we treated it as foundational social infrastructure? What if we approached AI as a kind of universal public service, designed to build collective civic agency and human flourishing?
Australia has the opportunity to build Universal Basic Intelligence (UBIQ).
UBIQ is not about subsidising access to AI models or a new entitlement. It is the commitment to build and maintain, within the population, a universal base layer of cognitive capability and agency that every Australian can access, contribute to, and use for their own ends, regardless of background. The deeper opportunity is to enable citizens to become co-intelligent.
A co-intelligent society is one in which human judgment, creativity, and local knowledge combine with machine capabilities to expand what communities can achieve together. Just as literacy transformed people’s ability to participate in society, co-intelligence transforms people’s ability to think, create, learn and contribute.
Building on compute and income: The next evolution
AI equity discussions tend to focus on two rising ideas: Universal Basic Compute (UBC) and Universal Basic Income (UBI).
Both concepts are relevant. Ensuring access to AI tools (compute) is a necessary baseline, and guaranteed income support may well become an important floor for economic dignity in the years ahead as AI impacts the nature of work. But neither goes far enough.
Universal basic compute gives communities the tools, but even guaranteed access does not guarantee that citizens know how to use, design, and harness the technology for their own ends. Universal basic income gives financial security, but financial support does not grant citizens the agency to shape the knowledge and cognition that could define their future.
UBIQ builds on these foundations. Grounded in Australia’s social-democratic heritage, Universal Basic Intelligence recognises that participating in, contributing to, and directing intelligence is both an individual right and a collective civic responsibility. It is an idea that sits in the public domain rather than the market, and it is enabled by, but is not purely within the domain of government.
The public school metaphor: Principles for UBIQ
Australia already has a model for the common good that builds human capital while reinforcing local identity: our universal public school system.
Public schools do much more than teach literacy and numeracy. At their best, they are foundational anchors of Australian civic life. They are where Australian stories are written, preserved, and shared:
- Open to everyone, irrespective of means: They establish a universal floor. Families can opt for private schooling if they choose, but a high-quality public option belongs to every child and family. Everyone is included.
- Agnostic to ends, committed to agency: Public schools do not dictate the paths students’ lives should take. They build agency, judgment, and critical thinking, allowing individuals and communities to define their own futures.
- A place for joy, as well as learning: Public schools are not only academic spaces. They exist for art, sport, play, socialising, and community engagement, all of which are key parts of the human experience.
- Trusted, local community anchors: They don’t just serve students, they serve their local communities. They are venues for community meetings, voting, and disaster hubs, built on deep-seated trust.
- Preserving local and Indigenous wisdom: They are spaces where First Nations knowledge and environmental empathy are passed down, alongside the diverse stories of successive generations of migrants who have made Australia home.
- Alive to equity and power: A great public school meets communities where they are, deploying resources to lift the outcomes for students with disabilities, cultural and linguistic diversity, and those experiencing disadvantage.
What if our national approach to AI had all these characteristics? What if we imagined our engagement with AI as a social infrastructure that respects, holds, and elevates Australian knowledge, while overtly empowering our citizens of today and tomorrow?
Moving from access to agency
Closing the digital divide in Australia has tended to focus on physical access: laying fibre-optic NBN cables, deploying outback satellites, distributing laptops to school kids, subsidising internet access and connecting remote Indigenous communities.
While access remains vital, it is not the only challenge. Some types of AI are already largely free (and rapidly improving in quality). While there are real sovereignty risks that governments must manage to ensure access, taking a productivity or economic inclusion lens is too narrow a frame. We don’t need the same computing infrastructure to underpin meteorology or genome sequencing as we do to run public schools; the same goes for the AI infrastructure needed for UBIQ.
The real challenge for Australia is agency. Just as a school is not merely a collection of books and buildings, AI agency is formed through people, relationships and doing – not through material things. True agency is not built through government programmes and online courses. Citizens need the ability to critically evaluate, challenge, and shape AI for their own ends. Agency is therefore not imposed or even induced. It is built through relationships, peer learning, and co-creation within local communities in neighbourhood houses, cafes, community gardens, public libraries, and First Nations organisations.
In the same way that literacy expanded participation in society, the ability to collaborate effectively with AI expands an individual’s capacity to learn, create, understand and act towards their own goals. Agency is increasingly a product of co-intelligence.
What would it take to make AI a universal public service?
If we are serious about treating AI like our public education system, what would need to be true?
- Distributed community capability: AI capability cannot be contingent on having the right job or provided centrally by government departments. It must be embedded as a trusted, accessible, designable resource in local community hubs across urban, regional, and remote Australia, supported and sustained by people in communities for their benefit.
- Long-term public commitment: Just as we do not shut down public schools during a downturn in the budget cycle, a universal public service requires stable, long-term funding and non-partisan support across the political spectrum.
- A diversity of local approaches: There is no “one-size-fits-all” approach. A local initiative in regional Western Australia will look very different from one in urban Western Sydney or an Indigenous community in Queensland. Capacity must respond to local communities and conditions.
- Cultural and sovereign grounding: Public education relies on a curriculum that reflects Australian history, values, and our diverse cultures. Similarly, a public AI base layer must be trained on and respectful of local knowledge systems, including tacit knowledge, Indigenous ways of knowing and migrant perspectives, rather than merely reflecting dominant overseas worldviews. This goes beyond training models on Australian data; it’s about ensuring that non-digitised forms of Australian knowledge are represented on our UBIQ, and that citizens build, not just use, AI.
AI leadership for human flourishing
Australia is well placed to lead on UBIQ. We combine strong social-democratic foundations and trusted public institutions with a vibrant, multicultural civic culture. We do not share the techno-scepticism of some countries, and we still have a sufficient footprint of civic institutions to act as the bedrock for community capability.
We do not have to settle for being a sub-market for US commercial platforms, an adopter of European regulation, or a bystander in Chinese coalition-building. By framing AI as a universal public service that builds—and builds on—community capability, Australia can offer a practical, human-centred template for other middle-income, middle-power nations across our region and around the world.
Universal Basic Intelligence is not an end in itself. It is the foundation upon which a co-intelligent society can be built. The opportunity before Australia is not merely to increase access to AI, but to expand the capacity of communities to think, learn and act together.
It is an invitation to ask a fundamentally different set of questions about our collective future:
- How do we shift power and agency to the communities who need it most?
- How can government act as an enabler and funder of community capability, rather than a top-down provider?
- And how do we ensure that the next generation of intelligence remains deeply human, local, and democratic?
Over the next three articles in this series, we will explore who risks being left behind in the current rush, how human wisdom and machine intelligence can partner through co-intelligence, and how we can practically build this capability from the ground up.
