Book Review: The Artificial Inventor - A Challenge for the Patent System

Image: Thomson Reuters
In ‘The Artificial Inventor’ (Thomson Reuters), Luz Sánchez García (University
of Murcia) characterises humanity as standing at the cusp of an ‘Artificial Invention Age’
in which Artificial Intelligence (AI) is no longer used as a tool but rather a creative partner or independent innovator. The challenges presented by this state of affairs, including whether ‘artificial agents’ can be considered inventors, patent their inventions, and enjoy the benefit of patent ownership – and how their inventions would be evaluated against established human rules in the first place, are considered in this book.

Suggesting the adoption of proactive legal solutions in order to “get ahead of the immediate future” (p.25), Sánchez García emphasises the importance of centering human rights. This approach would be intended to safeguard the interests of humanity at large as well as the societal interest in innovation, serving the purposes of patent law. 

The first of the book’s four chapters addresses the concept of AI, seeking to define and review the discipline. Kat readers active in this field and those who have been following criticism of the European Commission’s proposed AI Act will be well aware of the drawbacks of such exercises (see previous Kat coverage), but the author presents a useful overview of the potential benefits and pitfalls of innovation in this area, economically and in terms of global competition. The way in which the book suggests legal changes which could involve quite a fundamental reorientation of attitudes towards innovation and incentives is supported by a sweeping overview of millennia of human-machine history leading up to our present moment, in what is characterised as a ‘third wave’ of AI: “a new golden stage of boom and growth for AI” (p.37, n.50). This overview includes consideration of various different ways of categorising AI: the cognitive model approach, the Turing Test approach, the approach to the Laws of Thought, and the rational agent approach, before moving to a “comparative panorama of the robotics sector” (p.49). Sánchez García accordingly engages with the issues on a global level, as befits her description of the fundamental challenges presented – though does ultimately in this chapter pay the most attention to developments at European level, given that this is the region in which the most concerted efforts to date have taken place to legislate specifically for AI and to implement supporting initiatives.

The second chapter moves to the notion of an ‘Artificial Intelligent Agent’, including whether such an agent could possess both intelligence and consciousness. Assuming that consciousness is not at play, how then should AIAs be characterised legally? As with the overviews of various concepts in the previous chapter, Sánchez García moves to consider a range of potential options, from personhood, company status, animals, exceptional subjects of law, or slaves, with a particular focus on Spanish law but incorporating insights from other jurisdictions and historical developments. A more extended consideration of ‘electronic personality’ is followed by an alternative proposed definition, taking into account potential technological developments in the degree of autonomy of AIAs, that of the ‘performance imputation center’: a tailormade solution as an entity lacking legal capacity but capable of being attributed the ability to obtain an invention.

Keeping this backdrop and suggestion in mind, the third chapter addresses the question of whether a machine can “become creative” (p.97). Indeed, creativity and originality are described as already being manifested in the output of some AI solutions; it is even possible that technological singularity might be achieved in the coming decades, with DABUS receiving an honourable mention here (for more on this device, see previous Kat coverage). This question is addressed through the lens of general theories behind intellectual property, namely utilitarianism, the labour theory of property, personality theory, and the social planning theory. Again with insights gleaned especially from the Spanish experience, the author suggests that innovation is of such high importance that inventions generated by an AIA are worthy of patent protection, even if that AIA cannot itself respond to economic incentives. Approaching the concepts of author and inventor, Sánchez García prefers to adopt a new concept, the ‘Artificial Inventor’ as the “thing that produces effects”, capable of “influencing society with quantifiable and attributable effects” even if it cannot enjoy moral rights (p.121).

The final chapter presents a proposal for a legal regime with the addition of an “Invention’s Conception Contribution Test” (ICCT) in order to evaluate the AIA’s level of involvement in the conception of the invention. Subsequently, an assessment of whether ownership of the inventions to that agent could be attributed would be necessary. This would be supported by an AIA Registry and a fund for their inventions, within the framework of a ‘human control approach’. The ICCT is a novel contribution which is inspired by elements of the US’s former Interference Proceeding to adjudicate in cases of conflicting claims to priority. This chapter thus considers a range of issues from procedural elements to patentability, especially the ‘inventive step’ and the ‘person skilled in the art’ concepts – suggesting the adoption of an ‘expert system’ approach for application to the AIA situation. Ownership then remains the most significant outstanding issue at this stage of the book, resolved by the suggestion of a threefold system which distinguishes between ownership by the AIA creator, AIA user, and AIA itself, and buttressed by a suggested AIA Registry to achieve transparency and aid in dispute resolution. Lastly, a fund for the management of inventions generated by AIAs is suggested as a pool from which e.g., damages could be paid as well as conservation, repair or improvement costs (p.158). This could also aid transparency through being segmented according to the specific invention, in cases where one AIA generates multiple inventions. 

One small comment this Kat would make is that the punctuation in the English translation appears to follow Spanish norms, as does some of the phrasing. However, these did not affect the argumentation or information conveyed in this helpful primer to new and developing AI technologies – and how best the law should (re)act. In addition to making comprehensive suggestions for how the law, especially at European level, might do so, this book would be of interest to all who seek an introduction to the history and concepts underlying the present AI debate.

Publisher: Thomson Reuters

Publication Date: 2021

Extent: 188 pages, including notes and bibliography

ISBN: 978-84-1390-904-2

Book Review: The Artificial Inventor - A Challenge for the Patent System Book Review: The Artificial Inventor - A Challenge for the Patent System Reviewed by Sophie Corke on Monday, October 10, 2022 Rating: 5

1 comment:

  1. I have read the book and my conclusion is quite simple: a nice playground for legal scholars and a lot of conclusions based on the hope that one day machines will be better than humans which I cannot subscribe whatsoever.
    The book is interesting only in that it shows how far the fantasy of people not directly involved in technology and its applications come to discuss matters for which they have a rather limited knowledge.


All comments must be moderated by a member of the IPKat team before they appear on the blog. Comments will not be allowed if the contravene the IPKat policy that readers' comments should not be obscene or defamatory; they should not consist of ad hominem attacks on members of the blog team or other comment-posters and they should make a constructive contribution to the discussion of the post on which they purport to comment.

It is also the IPKat policy that comments should not be made completely anonymously, and users should use a consistent name or pseudonym (which should not itself be defamatory or obscene, or that of another real person), either in the "identity" field, or at the beginning of the comment. Current practice is to, however, allow a limited number of comments that contravene this policy, provided that the comment has a high degree of relevance and the comment chain does not become too difficult to follow.

Learn more here:

Powered by Blogger.