On 16 September 2025 the Full Court[1] delivered a fourth judgment on whether four of Aristocrat’s innovation patents are to a ‘manner of manufacture’,[2] finding that they are. The claims are to an electronic gaming machine (EGM) programmed in a particular way. Across the four decisions (trial, two Full Court and High Court) the tally is now 7 judges in favour of patentability, and 6 against.
Much of the commentary so far has suggested that this latest decision establishes a more favourable landscape for the registration of patents for computer-implemented inventions generally. This note questions whether that is the case, at least beyond EGMs, and considers what, if anything, has changed in terms of the applicable legal principles.
In summary, the judges who found in favour of patentability characterised the invention as a bespoke EGM that works in a new way, whereas the judges finding against patentability characterised the invention as a new gambling method provided by a form of computer (an EGM).
I have assumed that the reader is familiar with the history of this proceeding. I also pass over the 2nd Full Court’s reasoning on why the High Court’s decision was not precedential, and why the 2nd Full Court felt free to reverse the 1st Full Court for all claims other than the representative claim. The reasoning of the 2nd Full Court on manner of manufacture issues is at [126]-[134].
An artificially created state of affairs and a useful result
The 2nd Full Court at [131] said that the required approach for assessing whether there is a manner of manufacture for a computer-implemented invention is:
to ask whether, properly characterised, the subject matter that is alleged to be patentable is: (i) an abstract idea which is manipulated on a computer; or (ii) an abstract idea which is implemented on a computer to produce an artificial state of affairs and a useful result. [Emphasis added]
This was adopted from the Gordon J group’s judgment, which in turn was derived from NRDC.[3] The requirement that there be an artificially created state of affairs and a useful result is nothing new. All 13 judges (the trial judge, six High Court judges, and now six Full Court judges) have proceeded on the basis that previous case law on computer-implemented inventions was correctly decided, including in IBM, CCOM, Grant, Research Affiliates, RPL Central, Encompass, Rokt, Repipe, UbiPark and Hytera. All of those cases cited and purported to apply NRDC, as did the 1st Full Court.
The conundrum is how to determine whether there is a relevant artificially created state of affairs, or adopting the 2nd Full Court’s language above, how to ‘properly characterise’ the claimed invention.
The conundrum for computer-implemented inventions
A computer-implemented invention will always involve an artificial effect – at a minimum there will be changes that take place within a computer’s memory. Moreover, if the claim is to a computer that has been programmed to operate in a particular way then the claim will be to a physical object, a machine. From either perspective there is subject matter that involves an artificially created state of affairs and so a prima facie manner of manufacture.
Some may advocate that this should be sufficient to establish patentability. But the practical consequence of such an approach is that patenting the computer-implemented method allows the effective monopolisation of that otherwise unpatentable method because often it will only be commercially useful if carried out on a computer.
Courts have rejected a conclusion that an unpatentable method becomes patentable subject matter when computer-implemented, but to do so they have had to resort to looking to what they assess as being the substance, not the form, of the invention. That then reframes the critical question as being how to properly characterise the invention.
A side note – mechanical EGMs as a false comparator
This is a convenient point to critique part of the buttressing reasoning by the 2nd Full Court, which raised an argument that also found favour with the Gordon J group in the High Court and with the trial judge. The 2nd Full Court said[4] that it would be anomalous to find that a ‘computerised EGM’ was unpatentable because it would be patentable if it operated mechanically.
As a preliminary matter, there was no evidence referenced that showed that the features of claim 1 could be achieved by a mechanical gaming machine. But if they could then that involves a method being deployed by a form of mechanical computer – see en.wikipedia.org/wiki/Mechanical_computer.
Exactly the same hypothetical comparator could be considered for any other case involving a computer-implemented invention – assume that it might be achievable by a mechanical computer, assume that anything mechanical is a manner of manufacture, and then reason that the electronic equivalent should similarly be patentable.
That argument is moreover unnecessary. The equivalent argument could be made more directly, based on the fact that an electronic computer is itself an artificially created product, and therefore is itself a manner of manufacture. But that is an argument that has been rejected by previous case law as elevating form over substance, with reasoning supported by all 13 judges in Aristocrat.
How is the substance of the invention to be assessed?
The 2nd Full Court noted[5] that all six High Court judges had agreed that it was important to first characterise the claimed invention, as a matter of substance, not form, and that this was consistent with past Full Court decisions. But, with respect, the 2nd Full Court has provided no real guidance as to how that characterisation is to be approached.
While the 2nd Full Court briefly summarised previous decisions in which courts had found there was no manner of manufacture for computer-implemented inventions and said that in each case there was no artificially created state of affairs, it did not articulate how it made that assessment, for which characterisation of the invention is essential. It did not seek to distil the principles that informed the characterisation in each case. It similarly briefly summarised previous decisions in which courts had found that there was a manner of manufacture for computer-implemented inventions.
The 2nd Full Court set out its preferred characterisation at [127] for the claimed invention, which adopted the Gordon J group’s characterisation in the High Court:[6] ‘an EGM incorporating an independent player interface and a game controller which includes feature games and configurable symbols’.
The ‘game controller’ was found by the 1st Full Court to be a computer, and a standard part of an EGM, as were the features allowing a player to interact with the machine (buttons, display area etc.). The new features of the EGM were the elements of the gambling method, achieved by the programming of the game controller. The ‘feature games’ related to awarding free or bonus games to a player, and the ‘configurable symbols’ related to assigning either values or other symbols for use in those feature games.
The 2nd Full Court appeared to be influenced by perceived similarities with the facts in CCOM which it summarised at greater length,[7] which found that a computer-implemented method of assembling Chinese characters based on their component strokes was a manner of manufacture. But that invention could also be distinguished because it relates to the broader operation of a computer, facilitating a diverse range of uses of that computer for a Chinese language speaker. Here, the EGM is described as having a relatively narrowly defined mode of operation.
The Court said that its preferred characterisation was supported by expert evidence that the configurable symbols provided a ‘significant enhancement of a player’s experience’. That evidence was not referred to by the primary judge, and although it was noted by the Gordon J group it did not seem to inform their reasoning on characterising the invention. The 2nd Full Court did not spell out what was significant about its characterisation, or why its characterisation demonstrated that there was a relevant artificially created state of affairs. It appears, though, that their characterisation is one that has the artificially created state of affairs as the EGM itself, programmed to have the generalised ‘feature games’ and ‘configurable symbols’. But what then is the significance of a player’s experience being enhanced? And was their analysis impacted by whether the ‘feature games’ and ‘configurable symbols’ were new or not? Is an EGM itself simply a manner of manufacture, leaving the other section 18 and section 40 grounds to be considered separately?
The 2nd Full Court appeared to consider that an EGM that has been programmed to operate in a particular way is a bespoke product, perhaps in contrast with an ordinary computer that would be expected to be used in more than one way. Such a conclusion chimes with the Court’s bolstering reasoning based on a mechanical EGM (addressed above). It also chimes with the conclusion of the Gordon J group, which the 2nd Full Court appeared to support, which found that an EGM that is programmed in a particular way is an ‘altered EGM involving an artificial state of affairs’.[8] That reasoning of the Gordon J group was also based upon an arguably flawed assumption that a game involving physical components is patentable subject matter, such that it would be illogical for an electronic equivalent not to be patentable.[9]
In contrast, the 1st Full Court, and the Kiefel CJ group, focused on the EGM being a form of computer, with the only differentiating elements of the claims being the (unpatentable) gambling method that was programmed into the game controller.
It seems likely that the consequence of this decision will be that EGMs having new programming may be more readily found to be patentable. Whether there is any broader application to computer-implemented inventions is far less clear, particularly given that all 13 Aristocrat judges endorsed previous case law as being correctly decided, and this decision does not suggest any substantively different approach. In the meantime there remains the chance that this case goes to the High Court for a second time (with an odd number of judges sitting!)
[1] Aristocrat Technologies Australia Pty Ltd v Commissioner of Patents [2025] FCAFC 131: www.jade.io/article/1153938 . Full citations for the other cases referenced in this article can be found in this decision.
[2] The requirement in s18(1A)(a) of the Patents Act 1990.
[3] At [122]. It can be seen from [115] that this language was derived from NRDC.
[4] At [134].
[5] At [126].
[6] At [149].
[7] See at [129], [133].
[8] At [154].
[9] At [118]-[119]. See further ‘Not a Game Changer: Aristocrat v Commissioner of Patents’, Smith, C., Vol 130 IP Forum, Dec 2022 at 18.