Full Federal Court says no to AI inventors

In the latest chapter of the DABUS saga, the Full Federal Court of Australia has determined that a device characterized as an artificial intelligence machine cannot be considered to be an ‘inventor’ in Commissioner of Patents v Thaler [2022] FCAFC 62.

Background

Dr Stephen Thaler is the applicant for Australian patent application no. 2019363177 naming an AI machine DABUS as the sole inventor. The Deputy Commissioner of Patents initially considered that an AI machine could not be considered an ‘inventor’ within the meaning of the Patents Act 1990. Dr Thaler successfully sought judicial review, with the Federal Court concluding that an inventor under the Patents Act can be an AI machine.

The origin of entitlement to the grant of a patent lies in human endeavour

On appeal, the Full Court looked to the statutory language, structure and history of the Patents Act, and the policy objectives underlying the legislative scheme.

Their Honours found that in each of the relevant provisions throughout the history of the Patents Act, the ability of a person to make an application for a patent was predicated upon the existence of an “actual inventor” from whom the entitlement to the patent was directly or indirectly derived. Accordingly, only a person with a legal personality could be the “actual inventor”. Referring to the legislative intent, the Court noted that there was no indication that the role of the inventor would change or that the established law relating to entitlement was intended to be altered.

In relation to section 15(1)(a) of the Patents Act, which provides that “a patent for an invention may only be granted to a person who is the inventor”, the Court referred to the ordinary English meaning of ‘inventor’ as being the person(s) responsible for making the invention, namely, “the person who makes or devises the process or product”.

The Court also considered the meaning of ‘inventor’ in the context of section 138(3)(d) (relating to challenges to the validity of a patent on the basis that the patent was obtained by fraud, false suggestion or misrepresentation) and section 59 (relating to challenges to entitlement), and the relevant case law. Their Honours concluded that the law relating to the entitlement of a person to the grant of a patent is premised upon an invention arising from the mind of a natural person or persons.

Their Honours also considered the High Court decisions in Myriad and NRDC in relation to patentable subject matter, noting that:

…The assumption in both cases was that human agency was required in the development of the invention in suit. That approach accords with the legislative history to which we have referred, namely that the origin of entitlement to the grant of a patent lies in human endeavour, which is rewarded by the grant of a limited term monopoly.

It was therefore concluded that the Patents Act requires the inventor to be a natural person. In this regard, the reference to “a person” in section 15(1)(a) is a natural person and does not have the extended definition supplied by the Acts Interpretation Act (to include a body politic or corporate as well as an individual) given that there is sufficient contrary intention in the Patents Act. The Court also found that something without a legal identity, such as an AI machine, cannot give effect to an assignment.

Interesting Observations

The Full Court commented on the intention of Dr Thaler to provoke debate as to the role of AI in Patent Law and policy considerations, noting that such debate is important and worthwhile and should be attended to with some urgency. However, the present case is only concerned with the proper construction of the relevant legislative provisions as discussed above. Their Honours cautioned against approaching the task of statutory construction by reference to what it might regard as desirable policy, imputing that policy to the legislation, and then characterising that as the purpose of the legislation.

The proposition that if DABUS is not accepted to be an inventor, no invention devised by an artificial intelligence system is capable of being granted a patent was rejected. It was noted that DABUS being the inventor of Dr Thaler’s application was taken at face value. However, the characterisation of a person as an inventor is a question of law and the question of whether Dr Thaler’s application has a human inventor remains undecided. Had this question been explored, it may have been necessary to consider what significance should be attributed to various matters including that Dr Thaler is the owner of the copyright in the DABUS source code and the computer on which DABUS operates, and that he is also responsible for the maintenance and running costs.

Final Thoughts

Whilst the decision may be appealed to the High Court (Australia’s apex court), Australia has joined the US and UK in rejecting the notion that an AI machine can be considered an inventor under patent law, at least for now. Although it seems that the discussions are far from over. Indeed, the DABUS saga around the world has sparked much debate about the current and possible future capabilities of AI and the question of inventorship. Should a person who is an inventor be redefined to include an AI machine? Would such a change encourage innovation and benefit society? Should consciousness be considered when defining an inventor with respect to AI? Something tells me these discussions will go on for some time.

BE(Hons) Mechatronics MIP FIPTA

Helen joined the intellectual property profession in 2009 because of her passion for innovation and cutting-edge technology. Since then, Helen has successfully obtained patents, trade marks and designs for many businesses in Australia and overseas in a large number of technology areas including machine learning and image classification, second-life battery systems, automation, smart devices, medical apparatus, manufacturing processes and equipment, subsea mining technology, automotive technology, audio signal processing, embedded software, and control systems. Helen is experienced in drafting patent specifications, prosecuting patent application in Australia and overseas, providing infringement and validity opinions, conducting searches and providing general IP advice. Since joining Phillips Ormonde Fitzpatrick in 2017, Helen has developed a further specialisation in artificial intelligence and software patents. In 2019, Helen completed AI Programming with Python, a future skills course from Royal Melbourne Institute of Technology, where she gained hands on experience building an image classifier using convolutional neural networks, which enabled her to further develop her understanding and appreciation for AI and Machine Learning algorithms.