Artwork created by generative AI (genAI) and other software could be eligible for copyright protection, after a seven-year legal battle ended with a Federal Court ruling that gaming giant Aristocrat can patent game features created by its software without human input.

The latest ruling – which follows years of appeals and counter-appeals between Aristrocrat and the Commissioner of Patents (CoP) – found that “computer-implemented invention [CII] is a manner of manufacture” as defined in section 18 of the Patents Act 1990.

Aristocrat is forceful about intellectual property protection, with CEO and managing director Trevor Croker recently calling it “an ideas an innovation company at heart…. committed to protecting the great work of our dedicated creative and technical teams.”

In 2018, the company was denied four patents for a feature of its gaming machines that includes ‘trigger games’, which are displayed under specific conditions and include new “configurable symbols” set by the software and displayed to the player.

New physical designs for machines can be patented, but computer programs generally can’t because the law considers them analogous to an “idea” that the computer is having – logic that the Commissioner of Patents originally used to deny Aristrocrat’s claim.

However, the gaming firm argued that the new games could be patented because they were novel – and that creation of the new symbols had in fact changed the existing game machine into a completely new machine, with different rules and operations.

Aristrocrat appealed the CoP ruling, which was reversed in 2020 by the Federal Court, reversed by the Full Federal Court in 2021, again reversed in 2022 when a split High Court couldn’t reach a united decision, and again reversed in the latest Full Federal Court decision.

“In the 21st century,” the justices wrote, “a law… that is designed to encourage invention and innovation, should not lead to a different conclusion where physical cogs, reels and motors are replaced by complex software and hardware that generate digital images.”

“It is difficult to see,” it continues, “why the development of an implementation of an electronic gaming machine that utilises the efficiencies of electronics technology, would be disqualified from patent eligibility, when the old-fashioned mechanical technology was not.”

Tweaking a nuanced patentability standard

The idea that the way something is created – its manner of manufacture – determines its legally protection dates back to the 1623 UK Statute of Monopolies, which fed a complex body of legal precedents broadly asserting that only human creations can be protected.

That fuelled centuries of protection for physical inventions, feeding a body of precedents that was reflected in a 2023 US Federal Court decision that images, text and videos created by genAI machines can’t be copyrighted because they lack “human authorship”.

Yet the distinction is more nuanced than that, thanks to years of development of both computer hardware and software systems that have refined those standards.

The rapid adoption of genAI has further challenged intellectual property law: unlike software platforms built as specific sequence of source code, genAI’s fluid and ever-changing nature is a moving target for regulators.

In 2024 the UK Government updated the rules by released guidelines for patent examiners – including examples – that ruled that “AI inventions” can be patented where “the task or process performed by an AI invention makes a technical contribution to the known art.”

Suitable ‘technical contributions’ include where an AI invention “embodies or performs a technical process” existing outside of a computer; contributes to solving a technical problem in or out of a computer; or is a “new way of operating a computer in a technical sense.”

Setting a new precedent for the AI era

Those guidelines promise to shake up an area of law that, as with the Aristocrat case, have seen repeated arguments and legal reversals as courts shape innovation policy in an era where genAI tools are creating content and coding applications at breakneck pace.

AI already writes around a third of all code produced at Microsoft, the company’s CEO Satya Nadella confirmed earlier this year, with other firms at similar levels and surveys suggesting that many developers write more than half their code using AI.

With OpenAI reportedly set to launch a social network built on AI-generated videos, AI agents reshaping businesses and Europol warning AI will generate 90 per cent of online content by next year, ideas of authorship are changing at breakneck pace.

In light of these changes the Aristocrat decision “marks a pivotal moment in the ongoing legal discourse surrounding patent eligibility for computer-invented inventions,” Madderns patent attorney Christopher Wilkinson wrote in analysing the latest court decision.

Clarity around computer-generated content requires a “comprehensive and holistic approach to patent eligibility,” he added, calling the Aristocrat decision “a precedent that will influence future cases involving computer-implemented inventions.”

The decision in Aristocrat’s favour could spur a rush by other companies to patent their AI and software-invented creations – although Lucy Hartland, special counsel with Spruson and Ferguson Lawyers, notes that yet another appeal is possible before the predecent sits.

“There is much in those reasons to encourage those seeking patent protection in Australia for computer-implemented inventions,” she said, although the CoP could launch another High Court appeal if the ruling fails to provide “sufficient clarity… in analogous cases.”