Book review: The Future of Copyright in the Age of Artificial Intelligence

Can artificial intelligence (AI) create things? Yes. AI has created movies such as Sunspring 2016, written novels like The Day a Computer Writes a Novel, and made art in the form of The Next Rembrandt. 

But-are those creations the same as human creations? Arguably not. For a start, all those AI creations involved an incredible amount of human input. And, as Aviv Gaon points out in The Future of Copyright in the Age of Artificial Intelligence

AI-authored writing may look like any other human text, but it has no ‘soul’. Even the most advanced AI system that is running on a robust program, learning from a large volume of data and providing excellent results, is merely ‘guessing’ using probabilities that a word is appropriate in any given sentence. AI systems do not ‘understand’

Therefore, we should consider carefully if AI creations should be worthy of copyright protection? And, if so, to what extent and what does this mean for the concept of authorship in copyright?  Gaon (Harry Radzyner Law School, Reichman University, Israel) considers these and other questions. 

The book is presented in three parts. The first part, Searching For Common Ground, provides three chapters on the history and general concept of AI. The first chapter sets out the key developments in AI and lays out the current state of the technology. Chapter two addresses the philosophical question of legal personhood for non-humans with a view to refuting the main objections to AI personhood. The last chapter in this section then outlines the various definitions and approaches to the term ‘artificial intelligence’, arguing for new terminology. 

Part two, AI-IP Theory, Gaon devotes four chapters to considering whether IP theories can support IP protection for AI. In particular, by considering how and when IP protection for computer programs evolved.  He offers a theoretical framework for AI and reflects   on  the impact of AI on author-centric concepts, such as romantic authorship and individualism. 

The third part of the book, New Vision for AI Authorship, brings together the concepts discussed in the first two parts in four further chapters. It considers the paths that AI authorship may take in the future, and explores possible candidates for authorship in the AI creation process (e.g., the programmer, the user, or the AI itself). Gaon also considers no-authorship possibilities as alternative rights models, such as ‘author in law’ and AI moral rights. This section also includes arguments around possible amendments to copyright law, both on a practical and theoretical level, for example, through copyright exceptions or adjustments to standards of originality. Neither of these is straightforward because, naturally, rightsholders and AI developers find themselves on opposing sides of the arguments. As Gaon illustrates:

On the one hand “the current legal regime renders the AI market accessible only to dominant AI developers like Apple, Facebook, Google…” On the other hand, AI data used to create may be protected by copyright or privacy rights and therefore the use of that content would be an infringement of the rights both at input and output. Moreover, “with the ability to mimic human expression and style, AI systems pose a risk to the human author’s market.”

As readers may know, the UK Government has announced that it will introduce a new text and data mining copyright exception in relation to AI [Katpost here]. However, the parameters of this exception are not yet decided. Gaon suggests a sensible solution to balancing the interests of AI developers with those of the rightsholders, which is to narrow the exception to only text, and data mining for AI-training purposes as follows: 

This exception could feature a distinction between using data for commercial purposes and other public data (including social causes). Royalties, for example, can be applied when data is used partly or wholly to develop algorithms for commercial purposes.

This book is essential reading for those engaging with the debate around the development of the new copyright text and data mining exception in the UK. It also provides a thorough and thoughtful discussion on the concept of authorship in AI generally that will be of interest to scholars and students researching AI. 

Publisher: Edward Elgar

Available in hardback and eBook 

ISBN: 978 1 83910 314 8 

Extent: 288 pp

Book review: The Future of Copyright in the Age of Artificial Intelligence Book review: The Future of Copyright in the Age of Artificial Intelligence Reviewed by Hayleigh Bosher on Tuesday, November 01, 2022 Rating: 5

1 comment:

  1. What is valid for copyright is valid mutatis mutandis for patents.
    An AI machine cannot be considered as inventor as well as an AI machine cannot be considered as an author.
    An AI machine is only doing what it has been told.
    Where is the originality or inventiveness?
    Learning data should be public, as well as the correlation algorithm.
    This is the only way to be able to trust the result of an AI machine.


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.