This is a first-look interpretation of the 2026–27 Federal Budget from Robotics Australia Group, written for members and the wider robotics community.
It is a considered review aimed specifically at developers, manufacturers, integrators, researchers, end users and investors who'll be looking to us to translate the Budget into what it might actually mean for their businesses.
A note before I get into it: many of the most significant measures in this Budget — particularly the R&D Tax Incentive reforms — are announcements of policy intent, not yet legislation. The ATO itself says of the R&DTI changes: "this measure is not yet law". The detail will be worked through in drafting and consultation between now and the proposed 1 July 2028 start date. So everything below should be read as "where the Government is heading", not "where the rules already are".
Click here for the formal budget response
With that caveat in place, let me cut to it.
The short version
There are some genuinely good things in this Budget for robotics. There are some genuinely concerning things. And there's one section — the R&D Tax Incentive — where the headline number looks generous but the proposed package underneath is more complicated, and most commentary hasn't yet read the fine print carefully enough.
The wins: Defence is now committing serious money to autonomous systems at scale. Australian venture capital is being given expanded scope to invest in growing companies. A new National Resilience and Science Council is being set up to coordinate the fragmented mess of Australian innovation funding. All real, all worth welcoming.
The concerns: the Industry Growth Program — the most accessible non-dilutive grant pathway for early-stage robotics SMEs — is currently paused to new applications, received no new funding in this Budget, and has no announced successor. Australia's Economic Accelerator, which bridged university research into industry, has had $800 million redirected away from it. There is still no sovereign procurement preference, meaning a pilot with a government department tends to stay a pilot. And the headline 4.5 percentage point R&D Tax Incentive rate increase is, on closer inspection, paired with a proposed removal of eligibility for "supporting" R&D activities — the work that surrounds core experimental activity. For a hardware industry that lives or dies on prototyping, test rigs and validation, that proposal has the potential to be a real problem, depending on how the consultation lands.
So: it's a Budget with a strong defence-shaped foundation and some welcome capital-side reforms, but for everything outside defence the bridge between research and industrial deployment is weaker than it was last year, not stronger. Two years on from the launch of Australia's National Robotics Strategy, the framework is there. The mechanisms to deliver on it are not yet matching the ambition.
The rest of this piece walks through the detail. Where I'm describing proposed measures still to be legislated, I've tried to be clear about that.
What's actually good
Defence is becoming a real customer for Australian autonomous systems. Up to $15 billion over the decade is committed to autonomous and uncrewed systems within the 2026 Integrated Investment Program. The Government has explicitly named the Australian-designed and built Ghost Bat alongside "large numbers of low-cost drones for deployment". This is the strongest signal Australian governments have ever sent that field robotics — air, ground, maritime — is a sovereign capability. If you're building anything that flies, drives, sails or swims itself, this matters. And it matters even if you're not in defence directly, because defence demand pulls supply chains, skills and adjacent civilian applications along with it.
Australian venture capital is being given more room to back hardware companies. From 1 July 2027, Australian VC funds (the VCLP and ESVCLP structures, for the tax-curious) will be able to stay invested in companies that grow much bigger than they could before. The asset cap on companies that VCLPs can invest in is rising from $250 million to $480 million, with matching expansions on the early-stage side. Plain English: Australian money can stick with you for longer. The pressure to take a premature offshore acquisition — often the only way to access enough capital to scale — should ease. There's also a return of loss carry-back from 1 July 2026 (companies can apply losses against earlier-year tax) and new loss refundability for genuinely young startups from 1 July 2028. Both are practical recognitions that hardware companies often need to absorb significant early losses before reaching scale.
A new National Resilience and Science Council is being established. It will coordinate priorities across roughly $15 billion per year of R&D and innovation funding. This is the structural fix to a problem that has hobbled Australian innovation policy for years: funding scattered across multiple portfolios, no one pulling it together, no consistent national signal. Robotics Australia Group will be working to ensure robotics is at the table when this Council sets its early priorities.
The $20,000 instant asset write-off becomes permanent from 1 July 2026. Useful if you're a system integrator or small business adopter buying automation equipment — you can immediately deduct purchases up to $20,000 each instead of depreciating them over years. Modest but real, and permanence matters because you can finally plan equipment purchases without year-on-year uncertainty about whether the rules will still be there.
The R&D Tax Incentive — the proposal to read carefully
This is the part most likely to land wrong if you only read the press release. And it's the part most worth flagging that we are talking about proposed changes, not legislated changes.
The Government has announced an intention to lift the core R&D Tax Incentive offset rate by 4.5 percentage points from 1 July 2028 (subject to legislation). So the headline number for most companies would move from a 43.5 per cent net benefit to roughly 48 per cent. Combined with a proposed reduction in the intensity threshold from 2 per cent to 1.5 per cent, and a lift in the refundable offset turnover threshold from $20 million to $50 million, this looks like a generosity package.
Then you read the fine print. The proposed package, taking effect together from 1 July 2028 if legislated as announced, includes three further changes:
First, eligibility for supporting R&D activities is proposed to be removed entirely. Only core experimental activities would qualify. For most robotics companies, the work surrounding the core experimental hypothesis — prototyping, test rig construction, software infrastructure, technician labour, hardware validation — is a substantial chunk of total eligible R&D spend. The proposed change would move all of that out of scope. The original Ambitious Australia review recommended a "deemed rate" for supporting activities (a fixed offset benefit relative to core spend) to reduce ambiguity. The Government has gone further than that recommendation, proposing outright removal.
Second, refundable offsets — the cash refunds that startups and early-stage scaleups rely on while pre-revenue — would only be available to companies less than 10 years old. If you're an established robotics SME still investing seriously in R&D, you'd lose access to refundability under the proposed rules. That cash refund mechanism is exactly the thing that has kept many of our members alive through technical risk periods.
Third, the minimum eligible R&D spend is proposed to rise from $20,000 to $50,000 unless the work is done through a registered Research Service Provider or a Cooperative Research Centre.
Now the part the Government doesn't lead with. Their own costings show the proposed package as a net fiscal saving of around $690 million over five years: payments to claimants drop by an estimated $1.6 billion, partly offset by an estimated $910 million drop in receipts. If the package is legislated as announced, the arithmetic indicates a structural retargeting funded by narrowing the base, not a generosity measure. RSM has published a worked example based on the proposed amendments showing a nine-year-old hardware company whose claim drops from $783,000 to $480,000 — roughly a 40 per cent cut. That kind of profile is common across our sector.
Independent analyses from BDO and others have also raised that removing supporting activities, as proposed, would move Australia out of step with the way the OECD defines R&D (the Frascati Manual, if you want to chase the rabbit hole). This isn't a small drafting choice. It's a substantive change in what Australia would count as research and development.
Robotics Australia Group is not opposing the policy intent. R&D support can absolutely be better targeted. But the way "core experimental activity" is ultimately defined between now and 1 July 2028 is going to matter enormously to our sector — because in hardware R&D, the boundary between core and supporting work is much fuzzier than it is in pure software development. The detail will be worked through in drafting and consultation, and we are going to be heavily involved in that. Which brings me to the ask later in this piece.
The commercialisation pathway has weakened
Two years ago, Australia's National Robotics Strategy named three vehicles to deliver it: the National Reconstruction Fund, the Industry Growth Program (IGP), and Defence's Advanced Strategic Capabilities Accelerator. This Budget continues the first and the third. The second is currently in a different state.
The Industry Growth Program is paused to new applications, according to the official DISR and business.gov.au pages. No new funding was allocated in this Budget, on top of a $102 million cut at the December 2025 mid-year update. Both the Advisory Service and the grant pathway are affected. No successor program has been announced.
To be clear about what is and isn't confirmed: the program has not been formally closed. Existing grants and committed funding continue. But for startups and small robotics businesses considering an application, the door is shut, and there is no public signal about when (or whether) it will reopen. For our sector, IGP was the most accessible non-dilutive grant program for taking a prototype across the valley of death. Its current paused state, without a clear successor, leaves a real hole. I was fortunate to see firsthand the amazing level of innovation from our Aussie startups and SMEs and how invested and supportive the IGP Advisers were in connecting them into ecosystems and helping to commercialise their solutions - so this is particularly frustrating for me personally.
Separately, Australia's Economic Accelerator — the equivalent program designed to bridge university research into industry — has had $800 million of uncommitted funding redirected. Same gap, university side.
The R&D Tax Incentive (whether reformed or not) operates on a different mechanism and timeframe. It rewards R&D that is already happening; it does not, on its own, carry a hardware prototype from lab maturity into industrial deployment. The absence of a successor commercialisation program is, for our sector, the most significant gap in this Budget.
What else is missing
No sovereign procurement preference. Government purchasing power was identified in the National Robotics Strategy as one of the most powerful levers available to grow domestic demand. The Budget contains no "Australian-first" signal for civilian government procurement of robotics and automation. Without it, our pilots with departments and agencies tend to stay pilots, the scaled contract goes overseas, and the Australian taxpayer ends up funding the development of technologies that get commercialised somewhere else.
Physical AI is funded at a fraction of international scale. The up to $70 million AI Accelerator through the Cooperative Research Centres program is welcome but small. Our productivity gains from AI will not come solely from systems that generate text or images. They will come from systems that act in the physical world. Robots. The next phase of the National AI Plan needs to position robotics as physical AI explicitly, not as a separate or secondary track.
No robotics dimension to skilled migration or workforce reform. The $85.2 million investment in faster migrant trades skills assessments is welcome but targeted at electrical, plumbing and construction trades. There is no equivalent pathway for robotics roles, and no dedicated incentive for mechatronics graduates. The National Robotics Strategy identified workforce capability as a foundational pillar. This Budget does not yet act on it.
No funding for shared robotics testbeds. Internationally, shared testbeds for autonomous systems are an established mechanism for accelerating commercialisation, reducing duplication, and improving safety assurance. The Budget makes no specific provision for testbed infrastructure, despite the success of existing Australian facilities and the explicit recommendations of the National Robotics Strategy.
What we're doing about it
Robotics Australia Group's priorities for the next phase of innovation policy are:
- A commercialisation pathway that succeeds (or replaces, or restores) the Industry Growth Program for robotics SMEs, recognising the unique capital and time requirements of hardware development.
- A sovereign procurement signal for civilian government adoption of Australian-developed robotics and automation.
- Constructive engagement with Treasury during the period to 1 July 2028 on the proposed definition of "core experimental activity" in the reformed R&D Tax Incentive, to ensure prototyping, test rigs and hardware validation are properly recognised in whatever ends up being legislated.
- Explicit inclusion of robotics as physical AI within the next phase of the National AI Plan.
- Workforce and skills measures that recognise robotics as a distinct occupational pathway.
- Investment in shared testbeds and research infrastructure.
We are already engaged with DISR on the next phase of innovation and physical AI policy. We will be engaging with the new National Resilience and Science Council once established, and directly with Treasury on the R&D Tax Incentive consultation as the detail is worked through.
What we'd ask of you
Three things, in order of how much they help us:
Send us R&D Tax Incentive case studies. If you can model what the proposed 2028 changes might mean for your company — what proportion of your current eligible R&D spend is core versus supporting, how the proposed removal of supporting activities would affect your claim if legislated as announced — please send it. De-identified is fine. Real numbers from real companies move the needle in submissions to Treasury and DISR in a way that abstract complaints don't. We will use this in our consultation engagement.
Tell us about IGP impact. If you had a planned IGP application, an active application in flight, or built a commercialisation plan that depended on IGP advice or funding, let us know where the current pause leaves you. We need to be able to point to specific affected companies when we argue for the door to reopen or a successor program to be established. Having seen this firsthand, often the independent commercialisation advice provided by the Advisers was more valuable than the potential grant funding – so keen to hear about the non-grant side of this puzzle too.
Share the formal statement. The shorter, more formal response is on our website. Sharing it on LinkedIn or with industry contacts helps build the public weight behind what we're advocating for.
You can reach me at general@roboausnet.com.au
Closing thought
There's a temptation, in a Budget like this one, to either celebrate the wins or catastrophise the gaps. Neither is the right read.
The honest read is that the National Robotics Strategy framework is still there, defence has become a genuine anchor customer, the venture capital reforms are real and welcome, and the new Council is the right structural fix. At the same time, the commercialisation pathway has weakened, the headline R&D Tax Incentive figures sit alongside proposed changes that will need careful scrutiny in consultation, and the productivity infrastructure of the 21st century — robotics — is still not being treated as the core economic enabler it should be.
Two years on from the launch of Australia's National Robotics Strategy, this Budget is the first real test of whether government investment matches the Strategy's ambition. The signal is mixed. The framework exists. The detail of what gets legislated in the coming year will determine whether the next Budget can build on this or has to start again.
We'll keep you posted as the consultation processes unfold.
Share the formal statement. The shorter, more formal response is on our website: https://www.roboausnet.com.au/publications
Thanks for your support,
Paul Mason
CEO – Robotics Australia Group
Australia’s peak body for robotics
