Australia was never going to win the race to build the next OpenAI. It has no foundational-model giant, no hyperscaler born on home soil, and a domestic market too small to brute-force a consumer app to global scale. So the country is making a different wager: instead of chasing the digital-only playbook that the United States dominates, it is pouring capital and policy into deep tech — semiconductors, quantum, defence, advanced manufacturing — where defensible intellectual property, not user growth, is the moat. And crucially, the state is not standing on the sidelines. It is showing up as investor, as customer, and as ecosystem architect. India, facing its own version of the same strategic problem, is running a strikingly similar play.
The pivot
The shift underway in Australian startup land is a deliberate move away from the software-and-marketplaces model toward businesses built on hard, defensible technology. The logic is straightforward. A consumer app can be cloned and outspent; a fabrication process, a quantum control system, or a defence-grade sensor cannot be replicated with a marketing budget. For a mid-sized economy, betting on patents and physics is a more durable strategy than betting on network effects it cannot fund at American scale.
What makes the Australian version distinctive is the role of government. Per Startup Genome, governments are increasingly acting as investors, customers and ecosystem architects in strategic, capital-intensive sectors — AI infrastructure, semiconductors, defence and quantum among them. That is precisely the posture Canberra has adopted. The state is not merely handing out grants; it is taking equity-style positions, committing to procure from local firms, and shaping the conditions in which deep-tech companies can survive their long pre-revenue years.
The most underappreciated lever is the research-and-development tax credit. In Australia, founders can now treat the R&D tax incentive as a genuine funding alternative — a way to recoup a meaningful share of qualifying research spend rather than diluting equity in a punishing early round. For capital-intensive, science-heavy startups burning cash years before a product ships, that refundable credit functions less like a tax perk and more like a recurring, non-dilutive cheque. It changes the maths of staying independent long enough to build something real.
The capital behind it
Behind the rhetoric sits actual money. According to Forbes Australia, citing Startup Genome, founders can tap a pool of roughly A$1 billion earmarked for deep tech within the much larger National Reconstruction Fund Corporation — a vehicle in the order of A$15 billion designed to underwrite the country’s industrial ambitions. This is the clearest expression of the state-as-investor shift: public capital deployed not as charity but as a co-investor alongside private money in sectors the market alone would under-fund.
That public money is not flowing into a vacuum. Australia has built a credible private deep-tech stack over the past decade. Venture firms such as Blackbird and Main Sequence have backed companies built on hard science, and incubators like Cicada Innovations have specialised in shepherding researchers across the gap between the lab and a fundable company. The pattern, as Forbes Australia documents, is that venture capital is increasingly concentrating around defensible IP rather than purely digital apps.
Syenta, a semiconductor startup, is the kind of company this ecosystem is built to produce: deep technical IP, advanced-manufacturing ambitions, and a value proposition rooted in capability that is hard to copy. It is the sort of bet that would struggle to find patient money in a market obsessed with quick software exits — and exactly the sort of bet the combination of NRFC capital, specialist VCs, incubators and R&D credits is designed to make viable. The thesis is coherent: stack non-dilutive credits, public co-investment and specialist private capital, and you can keep a hard-tech company alive through the years when it has nothing to sell but promise.
The hard truths
None of this guarantees success, and the strategy carries real risks that deserve to be named plainly.
- Capital gravity points west. Global venture money is being sucked toward American AI at a scale Australia cannot match. When a single US foundation-model round can dwarf an entire national fund, persuading limited partners and founders to back patient, hard-tech bets at home is a constant uphill argument.
- Deep tech is slow and expensive. Semiconductors, quantum and defence systems run on timelines measured in years and budgets measured in tens of millions. The capital intensity is brutal, and the gap between a promising lab result and a shippable, certifiable product is where many companies die. Public co-investment helps, but it cannot compress physics.
- Talent and scale-up gaps. Australia trains excellent researchers, but retaining them — and the experienced operators who know how to scale a hard-tech company from prototype to global supplier — remains a structural weakness. Too many promising founders and engineers still drift to larger markets where the capital, customers and exits are denser. A fund can seed companies; it cannot, on its own, manufacture the late-stage scale-up muscle the ecosystem still lacks.
The honest assessment is that Australia has assembled the early-stage scaffolding well but has not yet proven it can carry deep-tech companies all the way to global scale. The next test is not whether the country can fund startups — it now plainly can — but whether it can grow them up at home.
The India parallel
India is running a recognisably similar experiment, scaled to a far larger economy and an even sharper strategic motive: sovereign capability. The country wants the ability to design and build critical technology domestically rather than depend on foreign supply chains it cannot control.
The instruments echo Australia’s. India’s Research, Development and Innovation (RDI) scheme is conceived as state-backed patient capital for high-risk, long-horizon technology — exactly the gap private venture is reluctant to fill. The IndiaAI Mission, meanwhile, positions the state as both funder and customer of AI compute and capability, ensuring there is demand and infrastructure to anchor a domestic industry. In both cases, as in Australia, the government is acting as investor, customer and ecosystem architect rather than a passive grant-giver.
The shared logic is sovereign capability versus global competition. Neither country can out-spend the US on foundational AI, so each is choosing where to build defensible strength — and using public capital to de-risk the years private markets won’t fund. The two ecosystems also have genuinely different things to teach each other. Australia offers a tighter, more disciplined toolkit: refundable R&D credits as a non-dilutive funding line, and a specialist incubator-to-VC pipeline that knows how to commercialise hard science. India offers what Australia lacks — a vast domestic market and talent pool deep enough to support scale-up, the precise stage where Australian companies stall.
What both still have to prove is the same thing: that state capital and patient policy can do more than launch deep-tech companies — that they can carry them through the long, expensive middle and out the other side as globally competitive firms. The bet is sound. The execution, in Sydney and in Bengaluru alike, is where it will be won or lost.
