05 October 2021

DABUS Again Denied in the US and the UK, Part III – Implications for Australia


In both the US and the UK, patent offices have refused to allow applications filed by Dr Stephen Thaler to proceed, on the basis that the named inventor – an ‘AI’ machine dubbed DABUS – is not a human being.  In the first article in this series I looked at the US approach to the role of the inventor in patent law and practice, and at the recent decision of Judge Leonie M Brinkema in the United States District Court for the Eastern District of Virginia (‘EDVA’) upholding the USPTO’s decision.  In the second, I discussed the split decision of the Court of Appeal of England and Wales, which upheld (by a 2-1 majority) the decisions of the UK Intellectual Property Office (UKIPO) and the High Court.

In Australia, the Patent Office also refused to allow a corresponding application by Dr Thaler to proceed.  In contrast to the US and the UK, however, that decision was overturned by Justice Beach in the Federal Court.  The Commissioner of Patents has now appealed that ruling to a Full Bench of the Court (case no. VID496/2021).  In this article, I will be looking at the potential implications of the recent US and UK decision for the conduct and outcome of the appeal in Australia.

It should be said at the outset that the US law is very different to that of Australia, and it is therefore unlikely that anything in Judge Brinkema’s legal reasoning will be influential upon the Full Court.  It has also become apparent through the UK High Court and Court of Appeal decisions that while the UK law shares some similarities with the corresponding provisions of the Australian Patents Act 1990, there are also some significant differences.  Even so, there are aspects of the reasoning of Lord Justice Arnold in the Court of Appeal that the Australian appeals court may consider persuasive, and that could therefore influence the outcome here.

There are two key questions likely to be addressed in the appeal, both of which also arose in the UK, although only the first received substantive attention in the US.  These are:

  1. Can DABUS, as an ‘AI’ machine and not a human being, validly be named as an inventor on a patent application?
  2. Can Dr Thaler, not being (at his own insistence) the inventor, establish a proper legal basis for entitlement to the grant of patents on inventions said to be generated by DABUS?

I shall look at each of these questions in turn.

Can a Machine be an ‘Inventor’

As I explained in the first part of this series, US patent law and practice are very different animals from those of Australia and the UK when it comes to the role of the inventor.  The US Patent Act is very clear that an ‘inventor’ must be a human being.  As of the commencement of the America Invents Act – promulgated by Congress only a decade ago, in 2011 – 35 USC § 100(f) includes an express definition of ‘inventor’ as an ‘individual’.  Furthermore, inventors are required to take an active role in progressing a US patent application, by executing an oath or declaration confirming that they are ‘the original inventor or an original joint inventor of a claimed invention in the application’.  There is, therefore, simply no scope in the US for a non-human to be an ‘inventor’, or to comply with the requirement to execute the inventor’s oath or declaration.

In the UK, the Court of Appeal also found that an ‘inventor’ must be a human being.  Notwithstanding the 2-1 split in the final outcome, all three judges were in agreement that although the UK Patents Act 1977 does not expressly define ‘inventor’, there are nonetheless a number of provisions of the Act that necessarily require an ‘inventor’ to be a natural person.

When Justice Beach similarly considered the Australian Patents Act 1990 as a whole, he reached a different conclusion.  Having determined that the (undefined) term ‘inventor’ is an ‘agent noun’ that could, in ordinary usage, encompass a machine, he further found nothing in the Act that would force a contrary conclusion (see, e.g., Thaler v Commissioner of Patents [2021] FCA 879, at [10]).  Certainly the particular provisions of the UK Act that led the Court of Appeal to conclude that an ‘inventor’ must be a natural person do not have direct parallels in the Australian Act.

So it is probably fair to say that there is nothing in the reasoning of either the US or UK decisions that could have any influence on the Full Court in Australia.  But does that mean that the outcomes in those countries are irrelevant?

One of the issues raised by the Australian government’s Productivity Commission (PC), in its 2016 Review of Australia’s IP Arrangements (‘PC Report’), was that ‘Australia still has a materially greater propensity to grant patents when the [European Patent Office] does not’ (PC Report, p 224).  Now, the PC was addressing the standard of inventive step here, but it had a broader concern.  As a significant net importer of patented technology, granting patent rights imposes a greater cost on Australia than on some of our stronger IP-generating trading partners (see, e.g., PC Report, pp 13 and 214).  One reason for the PC recommending the inclusion of an object clause in the Patents Act was to put additional focus on the interests of users and the public, which might lie in fewer, rather than more, patents being issued – particularly to foreign entities.  The government accepted this recommendation, and legislated to add section 2A to the Act, stating that:

The object of this Act is to provide a patent system in Australia that promotes economic wellbeing through technological innovation and the transfer and dissemination of technology. In doing so, the patent system balances over time the interests of producers, owners and users of technology and the public.

Taking this into account, we might well ask what policy purpose is served by allowing patent applications to be filed in Australia, and patents potentially to be granted from those application, where corresponding applications would not be allowed to proceed in the US, the UK, and other countries?  And while Judge Brinkema in the EDVA may have expressly declined to entertain policy considerations (‘[t]he Supreme Court and Federal Circuit have explicitly held that policy considerations cannot overcome a statute’s plain language, and that “[m]atters of policy are for Congress, not the courts, to decide.”’, Mem. Op. pp 15-16), Justice Beach was very happy to engage with the object clause, and did so extensively, opining (at [124]) that ‘[i]n my view it is consistent with the object of the Act to construe the term “inventor” in a manner that promotes technological innovation and the publication and dissemination of such innovation by rewarding it, irrespective of whether the innovation is made by a human or not.’

The way Justice Beach approached the application of the object clause is, I would suggest, precisely the opposite of the way in which the PC would most likely have approached it, had it been faced with the question of ‘AI inventors’ in the course of its inquiry.  There is little doubt in my mind that the PC would have concluded that Australia should not expand its definition of ‘inventor’ to allow more patent applications to be filed, particularly in circumstances where our major trading partners had not done so.  The PC would have argued that to do so would fail to properly balance the interests of Australian consumers, i.e. ‘users of technology and the public’.

It is apparent that the word ‘inventor’ in the Australian Patents Act could be construed to include machines, as argued by Thaler, or to be limited to natural persons, as argued by the Commissioner, and that Justice Beach considered the object clause in section 2A to provide some guidance in determining which construction is to be preferred.  Is this, then, something that could be revisited on appeal in view of the fact that the US and the UK (and, for that matter, thus far the EPO) have gone a different way?

Can Thaler Establish Entitlement?

Whether or not DABUS can be considered an ‘inventor’ for the purposes of the Australian Patents Act, it will be of no avail to Dr Thaler if he is unable to establish that he is entitled to be granted patents for inventions that it creates.  Normally, title to an invention, and any associated patent rights, is transferred from an inventor to another party by way of some legal obligation and/or contractual agreement, such as an assignment or a contract of employment.  But even if DABUS is an ‘inventor’, it is still not a ‘person’, and is incapable of owning or transferring intellectual property rights.

I should say, at this point, that technically the question of entitlement does not directly arise in the Australian case, as it did in the UK.  Thaler’s Australian application was originally filed via the Patent Cooperation Treaty (PCT), and formally lapsed because the Patent Office determined that Thaler had failed to satisfy the requirement under regulation 3.2C(2)(aa) that he ‘provide the name of the inventor of the invention to which the application relates’.  He is not required to establish entitlement at the time of filing.  He will be required to assert entitlement before the application can be examined, and he must actually be legally entitled to the grant of any patent, otherwise it will be liable for revocation under paragraph 138(3)(a) of the Patents Act 1990.  Lack of entitlement is also a ground of pre-grant opposition under s 59(a)(i).

Despite not being a basis for the lapsing of Thaler’s application, the question of entitlement was the subject of extensive submissions at the hearing before Justice Beach, and of a significant portion of the judgment (paragraphs [156]-[200]).  The reason for this was that the Commissioner sought to establish that it would be inconsistent with section 15 of the Act – which deals with entitlement – for a machine to be an ‘inventor’ because (it was said) there would be no legal mechanism for title in the invention to be transferred to a (natural or legal) person to whom a patent might ultimately be granted.  Thaler, of course, sought to establish the opposite proposition.

Thaler succeeded in this respect on the basis that he was in possession of the invention, in circumstances in which he ‘derives title to the invention from the inventor’, under paragraph 15(1)(c) of the Act, ‘by reason of his possession of DABUS, his ownership of the copyright in DABUS’ source code, and his ownership and possession of the computer on which it resides’ (Thaler v Commissioner of Patents at [193]).  Justice Beach thus accepted that there was a prima facie legal basis for Thaler to claim entitlement.  In reaching this conclusion he referred, in part, to Blackstone's Commentaries on the Laws of England (Clarendon Press, 1766), Book II (which is available for free, short-term, e-borrowing from the Internet Archive).

This is where things get interesting, because in the UK Court of Appeal Arnold LJ also referred to Blackstone, reaching the contrary conclusion that ‘there is no rule of law that a new intangible produced by existing tangible property is the property of the owner of the tangible property’ (Thaler v Comptroller General of Patents Trade Marks And Designs [2021] EWCA Civ 1374, at [137]).  Arnold LJ focussed on the doctrine of accession, as discussed in Blackstone at pp 404-405, whereas Beach J relied upon pp 405-407, in which Blackstone discusses ‘occupancy’ of intangible property.  Arnold LJ certainly did not ignore this subsequent passage, observing (at [132]) that:

Although Blackstone was well aware of the existence of intangible property … he did not suggest that the rule of accession applied to intangible property produced by tangible property, probably because such a possibility did not occur to him. Indeed, after the passage cited above, he pointed out in paragraph 8 at pages 405-407 that, although in Roman law the doctrine of accession applied to something written (i.e. the ink) by a person on another person's paper or parchment, Roman law had no concept of copyright in a literary work, whereas this was recognised by English law as being the property of the author. He also referred to patents “by virtue whereof a temporary property becomes vested in the patentee”.

It seems to me that Arnold LJ is primarily concerned with the existence of a rule of law that would enable Thaler to secure exclusive possession of the intangible output of a tangible machine, as a basis for his entitlement to be granted a corresponding exclusive (patent) right.  Arnold LJ concludes, however, that the doctrine of accession does not apply, because it is founded in the concept of ‘dominion’ (at [131]), whereas ‘intangibles are non-rivalrous goods (meaning that consumption by one does not preclude simultaneous consumption by others)’ and thus ‘the new intangible is not susceptible of exclusive possession’ (at [133]).

Beach J, on the other hand, goes directly to possession.  In this, I think he may be mistaken.  Blackstone states that what we would recognise as intellectual property (he is primarily discussing copyright) ‘is more properly reducible to the head of occupancy than any other; since the right of occupancy itself is supposed by [John] Locke, and many others, to be founded on the personal labour of the occupant.  And this is the right, which an author may be supposed to have in his own original literary compositions…’ (paragraph 8, p 405, footnotes omitted). 

This seems problematic for Thaler’s argument.  If it is DABUS’ ‘labour’ that brings the intangible into existence, then this raises the question of how a machine – which is not a person, and therefore cannot possess property – could be an ‘occupant’ or, conversely, how Thaler could claim ‘occupancy’ when he was not the one to exert ‘personal labour’.  I think this takes us back to Arnold LJ, and the conclusion that the law as described by Blackstone simply does not contemplate exclusive possession of ‘a new intangible produced by existing tangible property’.

Of course, it does not much matter what I think.  What does matter, however, is that one of the UK’s most eminent intellectual property jurists has expressed the view, in a judgment for the Court of Appeal, that there is no rule of the common law providing for Thaler to take exclusive possession of an invention generated by DABUS, and thus to obtain entitlement to the grant of a patent.  This would appear to be in direct conflict with the conclusion reached by Justice Beach, and with the proposition that the scheme of section 15 of the Patents Act 1990, which provides an exhaustive list of the persons entitled to the grant of a patent, is consistent with an ‘inventor’ being anything other than a natural person.

That might be right, or it might be wrong.  But either way, the opinion of an English Court of Appeal judge of the standing of Arnold LJ seems like something that judges in the Federal Court of Australia would be hard-pressed to ignore.


If Justice Beach had not been in such a hurry to hand down his judgment in Thaler v Commissioner of Patents – if, as is more common in the Federal Court, he had taken a few months rather than just 28 days – the decision of the UK Court of Appeal would have beaten him to the punch.  In that case, it is somewhat difficult to imagine that he would have issued the same decision.  At the very least, he may have felt obliged to engage with the reasons of Arnold LJ, and justify reaching a different conclusion.  Possibly he would have called the parties back to make submissions on the relevance of the Court of Appeal decision to the Australian case.  And he might have decided the case differently.

There is also, in my view, some doubt as to whether construing ‘inventor’ to include machines (or anything else, other than a natural person) is consistent with the object clause in section 2A of the Patents Act, given that this was added in the context of the government’s response to the recent Productivity Commission report.  At first instance, the Commissioner argued simply that the object was already embodied in the words of the Act, and that this includes the term ‘inventor’ referring only to a natural person, as required by the scheme of section 15. 

In announcing the filing of the appeal, IP Australia has stated that ‘[t]he decision to appeal does not represent a policy position by the Australian Government on whether AI should or could ever be considered an inventor on a patent application.’  This suggests to me that the Commissioner may be disinclined to press any alternative policy based arguments on appeal, leaving this territory open for Thaler’s team to make largely unchallenged submissions on the supposed economic ‘benefits’ of recognising machine inventors, as they did at first instance.

In any case, the recent developments in the US and UK can only serve to make the Australian appeal more interesting.


Post a Comment

Copyright © 2014
Creative Commons License
The Patentology Blog by Dr Mark A Summerfield is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Australia License.