As we come towards the end of our series of ‘Can AI be recognised as an inventor?’ related insights, we look at the approach to identifying inventorship in patent applications. Could a Non-Human Contributor (NHC) be the alternative approach?
As the debate continues to fulminate about whether artificial intelligence (AI) can be validly named as an inventor we consider the options. Most countries that have addressed this question (through the recent patent applications by Thaler that names DABUS as an inventor) have declined to recognise that AI can be validly named as an inventor. The exceptions are South Africa and Australia, which have both recognised that AI can be named as an inventor (although the Australian decision is on appeal).
With the current explosion of new AI applications and the increasing ubiquity of AI in new technologies, a new approach to identifying inventorship in patent applications will need to be developed in order to account for situations where AI was involved in the creation of an invention. Of course, not every invention that uses AI will use it in a way that allows that AI scope to create anything, and so the question of AI contributions to an invention will not arise in many cases. However, AIs are increasingly being used to generate intellectual products that would not be feasible for a human to create.
This raises the question of whether it would be proper to name a human as a sole inventor of an invention to which they only contributed in a relatively minor way, when compared to an AI contribution. It would not be truthful or accurate to omit the role of AI in the creation of that invention. In fact, under US law, if a person were to claim sole inventive credit for an invention that was predominantly created by an AI, this would potentially give rise to serious questions about whether that person was complying with the duty of good faith and candor imposed on all patent applicants before the US Patent and Trademark Office. The patent offices of other countries place related obligations on patent applicants to be accurate and truthful about naming inventors. The failure to recognise the involvement of AI in the creation of an invention will therefore potentially lead to serious problems under national patent laws of various countries about the accuracy or truthfulness of naming only humans as the inventors in a patent application where AI played a major part in generating an intellectual product that is described and claimed in the patent application.
AIs are tools. They are sophisticated and complicated tools, but tools nonetheless. What sets them apart from say, a beverage dispenser or a printing press is that they create an intellectual product (i.e. a decision or a series of decisions), rather than a physical product. It’s this potential for creating an intellectual product that brings AI systems into proximity with intellectual property issues, such as patents, designs, and copyright.
It may be opportune to note that AIs are not the only possible non-human contributors to an intellectual product. What about the monkey that took a selfie with a photographer’s camera? Should there be some provision that would allow for the recognition of that monkey as a contributor to authorship of the copyright in that photograph? With the potential for intelligent animals, such as primates, elephants, and dolphins, to also contribute an intellectual product, it may be helpful to refer to a “non-human contributor” or NHC as an over-arching class of contributor, rather than just AIs.
It is easy to imagine that humans will not be the only source of intellectual products generated in the future. We therefore may need to address the fact that “inventor” has traditionally only referred to a human. Until there is a political will to revise the definition of inventor, which may be never, we need some way to also recognise the contribution that a non-human may make to an invention.
With this in mind, the author proposes that, rather than arguing the change of the human-dependent definition of “inventor”, a new class of contributor be recognised alongside human contributors of an invention. This new class may be called “non-human contributor” (NHC), for example. This new class would supplement the listing of a human inventor by also listing any other contributors to the invention that would not otherwise fall within the definition of ‘inventor’.
This may take the form of a modification of the Articles of the Patent Cooperation Treaty, for example. A PCT Request filed under the PCT may, for example, be modified to include a check box to indicate that an NHC has been involved in the creation of the invention. There may additionally be a field to provide identification information or a description of the NHC to allow it to be suitably uniquely identified.
The benefit of such a new class of contributor is to provide increased accuracy and truthfulness about the basis of creation of the invention. Each country may then choose to recognise (or not recognise) such a non-human contribution according to their own laws.
We may get to a point in the future where there is no human involvement in the creation of an invention (e.g. where the AI contributor is created by another AI, which was created by another AI, and so on), and so to identify a human as an inventor of such an invention is not accurate. However, we are probably not ready as a society to tackle that yet. Meanwhile, the naming of a human alongside an NHC for an invention that involved some input from both would allow the proportional contribution of each contributor to be sidestepped, just as the relevant proportional contributions of co-inventors is typically treated as irrelevant (as long as they are each non-zero).
This proposed recognition of an NHC as a new class of contributor may also raise new issues for discussion, which is welcomed. This proposal is not intended to be a fully-developed thesis. Rather, it is intended to stimulate thoughtful debate and assist in moving us forward as we adapt to these strange and exciting times.