Full Court makes its decision on non-human inventors of patents - Commissioner of Patents v Thaler  FCAFC 62
In our article “Artificial Intelligence and IP Ownership: will the IP created by AI meet its maker?” we discussed some of the issues which arise in the artificial intelligence sphere (AI) in relation to intellectual property (IP) ownership.
In particular, we noted the decision of the Federal Court in Thaler v Commissioner of Patents  FCA 879 (Primary Judgement) which considered whether an AI program could be listed as the inventor for the purposes of a patent for an invention under the Patents Act 1990 (Cth) (Act). In that decision, the Court held that it was possible for a patent to have an AI program listed as the inventor on a patent registration, notwithstanding that actual ownership cannot be vested in the AI.
At the time of writing that article, we noted the Primary Judgment was subject to appeal to the Full Court of the Federal Court. That appeal has now been determined.
The Full Court’s decision
On 13 April 2022, the Full Court of the Federal Court made its determination of the appeal (see Commissioner of Patents v Thaler  FCAFC 62). The Full Court unanimously overturned the decision of Beach J and concluded that AI, a non-human, could not be considered an inventor for the purposes of a patent. We have summarised the relevant aspects of the decision.
Issues on appeal
The Court was asked to consider whether an AI machine could be considered to be an ‘inventor’ within the meaning of that term in s15 of the Act.
Section 15(1) of the Act provides that a patent for an invention may only be granted to a person who:
(a) is the inventor;
(b) would, on the grant of a patent for the invention, be entitled to have the patent assigned to the person;
(c) derives title to the invention from the inventor or a person mentioned in paragraph (b); or
(d) is the legal representative of a deceased person mentioned in (a) (b) or (c).
There was no dispute that in order to be granted a patent as the inventor in their own right under s15(1)(a), the person applying needed to be a natural person.
However, the primary judge had held that each of the limbs under s15(1) were distinct alternatives to one another and that the inventor subsequently referred to in each of s15(1)(b), (c) and (d) did not need to be the same ‘human’ inventor referred to in s15(1)(a).
On that basis the primary judge held that the application made by Dr Thaler (listing an AI program as the inventor) fell within the alternative limbs of s15(1)(b) and s15(1)(c).
While Dr Thaler was not the ‘inventor’ in his own right, as a natural person under s15(1)(a) he would be entitled to have the rights assigned to him or derive title to the invention from the AI program which invented the invention upon the grant of the patent.
A key issue on appeal was to consider this position and whether the four classes of persons in paragraphs (a), (b), (c), (d) were distinct and separate sources of entitlement, or whether a legal relationship was required between the inventor (referred paragraph (a)) and the person entitled to the grant under ss15(1)(b), (c) or (d).
The Full Court on the appeal held the correct interpretation was a legal relationship must exist between the ‘human’ inventor in s15(1)(a) and the person entitled to the grant. As such, a person under ss 15(1)(b), (c) and (d) can only be entitled to the grant of a patent after they have received that entitlement from a human inventor who is otherwise entitled under s15(1)(a). 
Accordingly, the Full Court confirmed that the ‘inventor’ named in the patent application must be a natural person .
In providing its decision on the meaning of inventor for the purposes of s15(1), the Full Court considered the history of patents legislation, including a range of historical case law.
The historical cases considered by the Full Court emphasised the requirement that a patent application demonstrate the inventor’s role in conceding the invention. Moreover, the former Patents Act 1952 (Cth) (which was superseded by the current Act) predicated the ability of a person to apply for a patent upon the existence of an “actual inventor” – that is, a person with a legal personality.
Their Honours also referred to the Statute of Monopolies 1623 (UK) (which is foundational to the principles of the current Act) which highlights the relationship between rewarding ingenuity and acts of invention by the ”true and first inventor”.
It should be noted “inventor” is not defined in the Act. That fact, however, is immaterial, having regard to the historical objectives considered by the Court and discussed above.
It is clear the legislature had not made any statutory alteration which would change the role or meaning of the inventor as a natural person as has been the historical position at common law. Accordingly, and because the entitlement to a patent is founded in human ingenuity and labour, the Court relied upon this historical meaning in forming the view that an inventor must be a natural person.
The Full Court’s approach maintains the present stance in Australia in relation to the interaction between AI and the law. That is, human authorship is paramount.
What is made clear by this case is that issues which arise with AI and IP ownership are at the forefront of people’s minds and people remain unsettled with the implications of our current legislation when it applies to AI and new technologies. With AI and other emerging technologies growing rapidly, the decision by the Full Court highlights the essential issues which may need to be considered by the legislature in order to ensure appropriate management of AI and any IP rights.
At the time of writing, Dr Thaler has filed an application for special leave to appeal to the High Court of Australia. Should Dr Thaler be successful, we expect that there will be further interesting developments in this case.
Authors: Maja Samardzic & Jonathan Harris
Supporting partner: Jason Sprague
 Commissioner of Patents v Thaler  FCAFC 62 at  – .