Software. It’s not just running our gadgets anymore; it’s the lifeblood of nearly every industry and aspect of modern life. Yet, when it comes to protecting the intellectual property behind this crucial technology – through patents – Australia finds itself in a frustrating and frankly, embarrassing position. The current system is not just struggling to keep up; it’s actively stifling innovation, creating a minefield of uncertainty for developers, and risking Australia’s relevance in the global tech landscape.
A Brief History of Software Patents in Australia
Australia’s history with software patents is a textbook example of a legal system failing to adapt to rapid technological change. In the early days of computing, the Australian Patent Office’s (now called IP Australia) stance was clear: software was nothing more than abstract ideas – mathematical algorithms, which were deemed as not patentable subject matter under Australian law. The idea was that software was simply a set of instructions, and that it lacked the necessary “physical or tangible” effect, or “manner of manufacture” as the Patents Act 1990 required. This effectively slammed the door shut on any attempt to patent software, no matter how groundbreaking.
The tide began to turn, albeit slowly and with much hesitation, as the commercial importance of software grew. The legal system was forced to acknowledge that software was not just theoretical but also capable of generating significant real-world impacts. Cases like CCOM Pty Ltd v Jiejing Pty Ltd [1994] FCA 1429 marked a turning point. This case, involving a Chinese language computer input system, established that computer-implemented inventions could be patentable, provided they had a ‘manner of manufacture’ that was considered to be a practical application, more than just the mere implementation of an idea with computer code. This opened the door for software inventions to be seen as more than just abstract ideas.
Anne Fitzgerald and Scott Phillips’ article, “Patentability of Software in Australia: CCOM v Jiejing“, the authors write:
“The decision in CCOM v Jiejing confirms that most forms of software-related inventions are now patentable in Australia, as computer programs are usually created with the intention of producing some “end result which is … of utility in the field of economic endeavour.”
However, this initial step was not a clear victory for software innovators. Instead, it marked the beginning of a long and painful period of legal uncertainty. Even today, decades later, the judiciary and the Australian Patent Office continue to struggle to interpret and apply the “manner of manufacture” concept to software, resulting in a frustrating landscape of inconsistent decisions and an overall stifling effect on software development.
The Aristocrat Debacle: A Symptom of a Broken System
The lack of understanding of modern software is further highlighted by the now infamous case of Aristocrat Technologies Australia Pty Ltd v Commissioner of Patents [2020] FCAFC 177. This case involved an electronic gaming machine (pokie machine) invention that had some complex software that operated through a network that triggered a ‘jackpot’ feature. However, even though the invention was integrated within a machine, the courts still focused on the software itself, stating it only used a computer as a “tool.” The High Court concluded that the invention was not patentable as it was ultimately determined to be an abstract idea that lacked a manner of manufacture.
This case was a huge blow to the tech industry, and it highlighted the problematic and inadequate way that the Australian courts approach the issue of patentable subject matter, and how that is applied to software.
The Aristocrat case, more than any other, epitomises the fundamental flaw in the Australian system. Judges are attempting to shoehorn cutting-edge software innovations into a framework designed for mechanical inventions of the past. It’s hardly the fault of the judiciary, as they are forced to view software inventions throuh the lens through an archaic system that was primarily created for mechanical (and medical) systems.
The Current State: A Legal Minefield
Today, the situation remains a mess. While outright rejection of software patents is not as common as it was in the past, the “case-by-case” approach, coupled with a confusing interpretation of the “manner of manufacture” requirement, creates an environment of immense uncertainty.
The law continues to focus on proving some kind of ‘technical effect’ or ‘practical application’ that goes beyond simply implementing a program on a computer. What counts as a “technical effect” is vague and subject to varying interpretations, leading to unpredictable outcomes. Judges often struggle to distinguish between something that is purely an abstract idea and something that provides a genuine technological advancement. This subjective approach means that nearly every patent application is a gamble, with no guarantee of success, even if the invention is truly novel and groundbreaking.
The legal system’s bias towards tangible, physical inventions means that many truly innovative software concepts are often unjustly excluded from patent protection. The focus on a “physical effect” or a “tangible effect” is completely inappropriate to modern software inventions that are often built on mathematical models and algorithms that are primarily intangible. The focus on these outdated criteria is not suitable to determine the real value and inventive step of a piece of software that is commercially important.
The current system essentially treats software as a second-class invention, and as such it fails to recognise that software is just as important, if not more so than the mechanical inventions of the past.
Australia vs. the World: A Backward Step
The inadequacies of Australia’s software patents approach become even more apparent when compared to other developed nations. The United States, despite some recent court challenges, has historically taken a far more permissive approach, granting patents for a wide range of software innovations. This has undoubtedly contributed to their dominance in the technology sector. While not without its issues, the US system at least recognises the value of software as an economic and innovative driver. The UK, while having more strict criteria for patentability than the US, has, with its European counterparts, still adopted a more flexible and pragmatic approach, focusing on the “technical contribution” and allowing many more software inventions to obtain patent protection than Australia.
Australia, in stark contrast, seems to be actively trying to hold back the tide of technological progress. Its overly restrictive and ill-informed approach to software patents has painted a picture of an Australian patent system that is simply not fit for purpose, and is actively out of touch with the rest of the world.
Why This Matters: The AI Catastrophe
The most alarming consequence of Australia’s outdated patent system, particularly for software, is the likely damage it will inflict on artificial intelligence (AI) innovation within Australia. AI is not some futuristic fantasy anymore, it is rapidly transforming industries and will continue to shape the future. Australia’s failure to properly protect the intellectual property that underpins these AI advancements will have devastating results.
AI is inherently software-based. The cutting-edge algorithms, deep learning models, and neural networks that drive AI are complex, often highly inventive pieces of software. Under Australia’s current software patents system, such inventions in the future are likely to face an uphill battle for patent protection, which effectively means that Australian AI developers are working at a significant disadvantage compared to their counterparts in the US or the UK, who have the opportunity to patent their inventions and seek to financially recover their investment.
The absence of a strong patent system for AI will discourage investment in the sector. Why would companies invest millions of dollars into AI research and development when their inventions can be freely copied by competitors? This lack of protection will stifle the growth of the industry, forcing Australian AI businesses to look elsewhere to seek protection and investment, therefore preventing growth within the Australian technology sector.
This lack of investment will undoubtedly lead to a significant skills shortage in Australia. Highly skilled AI researchers, knowing their inventions are not going to be protected will seek better opportunities overseas, where their work is recognised and rewarded. This results in a massive loss of intellectual property and expertise that Australia cannot afford to lose. Australia will therefore become a technology follower instead of a technology leader in the field of AI, which is now recognised as the primary technology of the future.
The consequences are clear: Australia could get left behind in the global software innovation race and it’s not only artificial intelligence,
As the world rapidly embraces AI, Australia’s inability to protect software effectively could see software developers take their operations offshore. The economic and strategic ramifications of this cannot be overstated. It will mean that Australia will have to rely on other countries to provide the technology to support its infrastructure which creates a massive dependence on other countries and potentially jeopardises our security. It also results in Australia’s failure to be able to compete in the global market and be able to fully participate in the rapidly evolving world.
The Need for Radical Change in 2025
Australia’s current patent system for software is not just inadequate, it is deeply harmful. It is a system that is stuck in the past, failing to understand the nature of modern software and its importance in a digital world. The system is actively discouraging innovation, investment and the development of a tech sector, particularly in the field of AI which will have lasting impacts for the future of the country.
Australia needs a fundamental overhaul of systems, which includes:
Politicians, through the Australian Parliament, must modernising the Patents Act 1990 (Cth): The “manner of manufacture” requirement needs to be either redefined or completely replaced with a test that takes into account the unique aspects of software inventions. The law needs to be brought into the 21st century.
Recognising the Value of Software: Software, including AI, should be treated as a legitimate field of invention, and not just a set of abstract ideas.
Providing Certainty: Clear, predictable rules are vital for businesses to invest in innovation. The current system creates uncertainty and risk, which undermines the system.
Consultation with the Tech Industry: Any reform should be developed in consultation with software developers, AI researchers and business leaders to better understand the practical challenges that they are facing.
The stakes are incredibly high. If Australia continues down this path, it will be actively choosing to fall behind in the global tech race and the world of AI. The current system is not only outdated; it’s a disaster waiting to happen. It’s time for bold, decisive action to create a patent system that encourages, rather than stifles, Australian innovation. It’s time to protect our inventors and future generations by providing a system that appropriately protects software in the modern world. We cannot afford to wait any longer.
