Data Reveals the Fragile Economics of Training Patent Attorneys in the Age of AI

Could the trans-Tasman (Australian and New Zealand) patent attorney profession be standing on a demographic precipice?  Despite the continued resilience of patent filings and a seemingly stable market for professional intellectual property services, the fundamental engine of the profession – its human capital – is barely idling.

I have analysed numbers over recent years from the Trans-Tasman IP Attorneys register, and the data reveals a profession that is effectively running on a treadmill.  In January 2018, the register recorded 1,003 attorneys.  Today, in mid-2026, that number sits at 1,084 (870 in Australia, 214 in New Zealand).

That is a net addition of just 81 practitioners over more than eight years, translating to a sluggish compound annual growth rate (CAGR) of just under 1%.  While the profession has managed to register 288 new attorneys since 2018, I found that this influx has been largely offset by a steady rate of retirement: 207 attorneys have departed the register over the same period.

On the surface, a 1% growth rate might simply look like a mature, saturated market.  But, when placed in historical context, the situation is concerning.  More importantly, it foreshadows a possible existential threat to the profession.  If the entry rate of new patent attorneys falls below the rate of senior attrition, the regulatory and commercial ecosystem may stand on shaky ground.  And – as always these days – the advancement of AI use within attorney firms threatens to exacerbate this risk.

 Read full article ► 

Tags: Australia, New Zealand, Patent attorney profession

Funding Quantum Hardware, Forsaking Research Talent: The Deepening Crisis in Australian Science

It has been just over two years since the Australian and Queensland governments announced their near-billion-dollar joint venture to fund PsiQuantum. Back in April 2024, policymakers excitedly pitched the construction of a utility-scale quantum computer at Brisbane Airport as an ‘anchor’ for the next industrial revolution. I did not write about it then.  Like many, I assumed the inevitable collision with technical and logistical reality would happen quietly, behind closed doors, perhaps eventually leading to a scaled-back scope or an extended timeline.

But here we are in mid-2026, and the wheels are visibly wobbling.  Last month, the much-touted Brisbane Airport site was unceremoniously abandoned in favour of a council-owned site at Moreton Bay Central.  Yet, despite pivoting to an empty lot at a former paper mill, it appears that neither the company nor the government has publicly walked back their foundational commitment to have the site operational by the end of 2027.  Promising to build the world's first million-qubit, fault-tolerant computer from the ground up in just 18 months is, to put it mildly, an ambitious logistical – and technical –  undertaking!

More importantly, I am writing about this now because we should not continue to ignore the sheer opportunity cost of this mega-project.  While the government plays venture capitalist with a single, highly speculative hardware gamble, the reality is that Australian science may be reaching a breaking point.

According to the Australian Bureau of Statistics (ABS), Australia’s Gross Expenditure on R&D (GERD) was a mere 1.69% of GDP in 2023-24.  This leaves us languishing far behind the OECD average, estimated at 2.93% in 2023, and significantly outpaced by peer economies that treat foundational research as critical infrastructure rather than a discretionary expense.

We are, quite literally, starving the ecosystem.

There is a profound irony at the heart of the PsiQuantum deal.  The company is headquartered in Silicon Valley, but its co-founders are products of the Australian university system.  They are the quintessential example of the ‘brain drain’ – talented innovators who left Australia when the domestic funding ecosystem failed to support their ambition.  This leaves us spending nearly a billion dollars of taxpayer money to effectively buy back our own exported talent at a massive premium.

Yet, in doing so, we are contributing to the collapse of the very STEM pipeline required to produce the next generation of innovators.  How does a government justify a near-billion-dollar bet on a speculative hardware project while simultaneously starving the foundation?  The answer, as is so often the case in technology policy, lies in a deliberate semantic conflation.  By using the word ‘quantum’ as a monolithic buzzword, policymakers have successfully blurred the lines between practical, deployable technologies and a long-term engineering marathon.

Let's dive in and look at the data.

 Read full article ► 

Tags: Australia, Funding, Government programs, Innovation policy, Manufacturing, Politics, Public policy, Quantum technology

Patent Filing Trends in 2025 Show Impact of AI and Continued Decline in Listed Firm Share

The Australian and New Zealand patent filing landscape in 2025 was marked by the continuation of established market trends alongside a striking new development that has potentially significant implications for both the patent system and the patent attorney profession.  Total standard patent applications filed in Australia decreased marginally by 0.4% to 30,327, representing the third consecutive year of modest decline following a pandemic-era peak.  New Zealand complete applications fell more substantially, by 4.4% to 5,538.  However, these figures were eclipsed by a dramatic surge in provisional patent filings driven by what appears to be widespread use of generative AI tools by self-represented applicants, a phenomenon that resulted in a near-tripling of self-filed provisional applications.

Market share dynamics continued to evolve in ways that present ongoing challenges for firms within the IPH Limited group.  Independent firms collectively surpassed the 50% filing share threshold for the first time, accounting for 51.7% of Australian patent filings, while IPH's share declined further to 32.7%, down from 35.0% in 2024.  Meanwhile, QANTM IP firms maintained relatively stable positioning at 15.6% filing share under private equity ownership.  This continuing divergence reinforces the conclusion that ownership structure alone cannot explain firm performance, with factors including geographic market exposure, client demographics, and strategic positioning all playing significant roles.

Analysis of the 2025 data reveals a market navigating multiple concurrent trends: declining US-originating applications (down 5.1%), growing Chinese filings (up 13.5%), and the emergence of AI as both a potential disruptor and, paradoxically, a possible source of future work for patent attorneys.

Let’s dive in and look at the data.

 Read full article ► 

Tags: Annual patent data, Australia, New Zealand, Patent analytics, Patent attorney profession

High Court Backs Aristocrat on Software Patentability – It’s Time for IP Australia to Follow Suit

The High Court of Australia has denied an application by the Commissioner of Patents for special leave to appeal a decision of the Full Federal Court in which claims to a computer-implemented invention (CII), in the form of an electronic gaming machine (EGM), were found to be patent-eligible in Australia.  This brings to an end the long-running dispute over patentability of Aristocrat’s claims, and marks the first time since 1991 (IBM v Commissioner of Patents [1991] FCA 625) that any such dispute with the Commissioner over claims to a CII has concluded with a victory to the applicant/patentee. 

The significance of this outcome cannot be overstated.  In recent years we have seen many cases conclude with claims being found ineligible – i.e. not for a ‘manner of manufacture’ under Australian law – namely Grant v Commissioner of Patents [2006] FCAFC 120, Research Affiliates LLC v Commissioner of Patents [2014] FCAFC 150, Commissioner of Patents v RPL Central Pty Ltd [2015] FCAFC 177, Watson v Commissioner of Patents [2019] FCA 1015 (leave to appeal refused in Watson v Commissioner of Patents [2020] FCAFC 86), Repipe v Commissioner of Patents [2019] FCA 1956, and Commissioner of Patents v Rokt Pte Ltd [2020] FCAFC 86.  That is a lot of cases telling us what is not patent-eligible, with absolutely no guidance from the courts on where the boundary might lie on the positive side of patent-eligibility.  Finally, we have a decision of the Full Federal Court that sets out a reasonably clear and concise test which is then applied to eligible claims.  The High Court, in denying the Commissioner’s application for special leave to appeal, stated that the ‘Full Court applied established principles concerning the assessment of manner of manufacture and reached a unanimous and clear conclusion as to characterisation’: Commissioner of Patents v Aristocrat Technologies Australia Pty Ltd [2026] HCADisp 15, at [2].

Following a rather convoluted passage through the Federal Court, the Full Federal Court and the High Court, back to the primary judge and then to the (differently constituted) Full Court, the final word on patent-eligibility of Aristocrat’s claims is Aristocrat Technologies Australia Pty Ltd v Commissioner of Patents [2025] FCAFC 131.  In this decision, the Full Court effectively overruled the reasons (though not the orders) of the earlier Full Court, and adopted the affirmative reasons of three High Court judges in the split decision Aristocrat Technologies Australia Pty Ltd v Commissioner of Patents [2022] HCA 29 (see High Court’s Failure Exposes the Festering Eligibility Sore in Australia’s Patent Laws for more).  Somewhat unusually, this also involved, in a sense, ‘overruling’ the negative reasons of the other High Court judges, in a rare instance of three judges of an inferior court asserting precedence over three judges of a superior court (albeit with the support of three further judges of that superior court).  But that’s just a function of how weird this case became!

In what follows, I will briefly summarise how we got here, where the law stands in view of the Full Court’s 2025 decision, and the High Court’s reasons for refusing special leave to appeal.  Finally, I will review the current examination practice of the Australian Patent Office, and the prospects of change in the wake of the High Court’s decision.

 Read full article ► 

Tags: Appeal, Australia, Computer programs, Examination, High Court, Patent Office, Patentable subject matter

Are Patent Offices Being Inundated with Low-Quality AI-Generated ‘Slopplications’?

‘AI slop’, defined as ‘low-quality content created by generative AI, often containing errors, and not requested by the user’, was named Word of the Year for 2025 by the Macquarie Dictionary.  Replace ‘content’ with ‘applications’, and ‘user’ with ‘patent office’ – let’s call them ‘AI slopplications’ – and we would have a good definition of a phenomenon that I suspect is occurring at offices around the world.  In the first 10 months of 2025 (i.e. up until the end of October) the number of provisional applications filed at IP Australia by self-represented applicants was up by a whopping 174% over the previous five years’ average! The overwhelming majority of self-filed applications (98.6%) originate in Australia (i.e. have at least one Australian-resident applicant).

The number of self-filed standard patent applications has also increased, being 82% higher in the first 10 months of 2025 compared with the previous five years’ average.  However, the number of standard patent filings by self-represented applicants remains a small proportion of the total – just 2.6% of all applications filed up until the end of October – and so the significance and impact of this increase remains to be seen.

The only plausible explanation I can think of for this sudden jump in filings by self-represented applicants after years of relative stability in numbers is the increasingly widespread and affordable availability of generative AI.  What is not yet apparent is how applicants are using AI.  Are they using ChatGPT and similar tools to assist in drafting patent specifications describing inventions made wholly by human inventors?  Or are they also using AI to facilitate invention itself?

Either way, I fear that this will not end well for many of these self-filers.  To be clear, there is absolutely no question that AI tools based on large language models (LLMs) can be used to assist in drafting patent specifications.  In the hands of an experienced patent professional who understands the invention to be protected, the full legal requirements for protection, and the various national and international drafting principles, even a general-purpose tool such as OpenAI’s ChatGPT, Anthropic’s Claude, or Google’s Gemini can accelerate the process of preparing a quality patent specification.  But there are also many potential pitfalls to using these tools, and they do not embody the significant expertise, skill and experience of a competent patent professional.  And if the AI is also contributing to the invention, then there may be nothing that is legally protectable at all!

The good news for Australian patent attorneys is that the increase in DIY (with AI) provisional applications has not been accompanied by a corresponding decrease in applicants engaging professional assistance.  The number of provisional applications filed via registered attorneys and firms for the first 10 months of 2025 is down by just 3.2% on the past five years’ average.  And while this does reflect an ongoing decline over recent years, it indicates that the use of AI may be bringing new users to the patent system, rather than taking work from professional advisors.

 Read full article ► 

Tags: Artificial intelligence, Australia, Large Language Models, Patent analytics, Patent drafting, Patent Office, Patents and society, Provisional applications