Abstract

The law is a stable regulator of public relations, but at the same time, it is very flexible, because it keeps up with the times and is changed, supplemented, and specified. In the age of technology development, which permeates all types of life activities of individuals, it is impossible not to note the growing importance of regulating legal relations in cyberspace, starting from the moment of entering this space. An important area of regulation of these legal relations is the mechanism for protecting intellectual property rights that arise in neuro-space. The field of intellectual property law is not so new, but its protection still needs to be developed. It is impossible to neglect the protection of intellectual property rights arising in neuro-space and be content with the presence of regulation of legal relations outside the digital space. Intellectual property rights, although newer than the standard list of rights, they fully claim to be attributed to an independent branch of law. The article deals with the main issues related to the copyright of Artificial Intelligence to create works of art: who is the owner of personal non-property rights, and who are the exclusive rights to the work.

  1. Introduction

A wide range of productions in the literary, scientific and artistic domains is protected by copyright. This not only includes cultural creations, such as works of literature, music, drama, film, photography and art, but also functional types of subject matter, such as computer programs, databases, industrial design and works of applied art. As a rule, copyright protects works regardless of their ‘merit’ or purpose: the design of ordinary household items is eligible for copyright protection just as much as creations of ‘high’ art. The only threshold that must be satisfied for a work to attract copyright is that its expression is sufficiently ‘original’, in the legal sense of that word. With the development and popularisation of the digital space, Artificial Intelligence began to develop with great steps and entered into all spheres of our life. To illustrate, Artificial Intelligence has “learned” to paint pictures. This, for example, is Next Rembrandt,[1] a neural network created jointly by ING and Microsoft. According to the official website, it analyses the artist’s works, collects information from them, recognizes the features of the letter and brings to life new works written in the style of Rembrandt.

There are many equally interesting examples of the use of Artificial Intelligence in art, but the main question that lawyers pose is who owns the copyright? Disputes around authorship arise only because of the desire of the copyright holders or creators of a computer program to receive material benefits. Different jurisdictions have their own solutions to this issue.

  1. Establishing the right holder

The creative powers of advanced Artificial Intelligence systems have led some scholars to conclude that the results of artificial creation cannot be protected by copyright, since human beings have lost control of the creative process.[2] Some writers, therefore, argue for the introduction of special neighbouring rights to protect “authorless” Artificial Intelligence-generated output against misappropriation.[3] But is this assumption correct? Or can Artificial Intelligence-assisted output qualify for copyright protection, despite the increasingly important role that machines play in its creation?

To compare, the Copyright, Designs and Patent Act of the Parliament of the United Kingdom of 11/15/19883 states that: “the author of a literary, dramatic, musical or artistic work, if it is created on a computer, is recognized as the person who takes measures to create this works”. But in some programs, for example, Adobe Photoshop, there is a function of automatic image enhancement (auto contrast, etc.). Does this mean that the creator of this mechanism also receives the rights to all photos edited in this way? Certainly not, because he does not even know which images are to be improved by his program. That’s why V.O. Kalyatin recognizes such a legal position as “clearly unsatisfactory”.[4]

The purpose of the institution of establishing the right holder is to support and encourage authors with property benefits. The computer program does not require funding and protection from illegal copying. The fact is that in this matter only, as already mentioned, the copyright holders of the program or its creators want to benefit. But they are already getting their benefit; this is, firstly, the cost of the exclusive right to the program itself, the possibility of issuing licences, the proceeds from the sale of the picture; secondly, market benefit (it is known that the value of shares varies depending on the introduction of innovations. In the example of NextRembrandt mentioned above, the capitalization of both companies increased by 20% for Microsoft and by 61.29% for ING).[5]

So, obviously in 2018 put up for auction Christie’s “Portrait of Edmond Bellamy”, completely created by a neural network. The company gained 432.5 thousand dollars from its sale. Having learned about the deal, the creator of the Barratt algorithm demanded that his name be indicated using the data of the operating time.[6] It turns out that there is no need to establish AI authorship because property issues are resolved at the level of persons participating in the creation of a computer program.

A similar point of view is also defended by V.O. Kalyatin in the above-mentioned scientific article. He writes that copyright protection of works is necessary for terms of remuneration. Such a defence of a right is not established merely from the fact that something has been created; it has an encouraging meaning. When answering the main question of the work, it is necessary to take into account T. Butler’s idea of ​​fictitious authorship.[7] According to this concept, the distribution of rights takes into account the contribution of everyone (company, specific programmer and other possible persons).  As a result, the court decides who owns the entire scope of rights:

  1. To the most “worthy” of all the persons represented;
  2. To all of them as co-authors;
  3. The right is divided into parts according to the principle “to each his own”;

This design was rejected because it represents too long a procedure for establishing authorship. However, why not establish a presumption of the right holder of a computer program? Its creator is always known. Like any author, his rights arise initially, but he can transfer them, for example, to a company under civil law contracts.

By the way, the construction of civil law contracts, for example, an author’s order contract, is permissible with the transfer of the exclusive right to a work of literature, science, and art. In this case, personal non-property rights are assigned to the author, and property rights to the owner of exclusive rights.

So, it seems possible to split the scope of copyright in matters of Artificial Intelligence. Exclusive rights and the right to judicial protection of the name of the “author” (i.e. Artificial Intelligence) should be left to the copyright holder of the program. The meaning lies in the fact that the copyright holder has a greater interest in protection than the neural network. Illegal copying of an image can lead to a decrease in the market value of the original and a decrease in the company’s capitalization.

However, since the figure of the creator of a work of art must exist, in order to protect the creation from appropriation, it is necessary to preserve the fictitious right of the author to the name for the neural network. Therefore, it will still be necessary to indicate the name of the Artificial Intelligence ​​for any operations with the picture. Otherwise, the work without the author does not create any copyright, including exclusive. This is confirmed in the third edition of the Compendium of U.S. Copyright Office Practices.[8]

  1. The copyright analysis

As our inquiry into EU copyright law reveals, four interrelated criteria are to be met for Artificial Intelligence-assisted output to qualify as a protected “work”: the output is (1) in relation to “production in the literary, scientific or artistic domain”; (2) the product of human intellectual effort; and (3) the result of creative choices that are (4) “expressed” in the output. Whether the first step is established EU law is uncertain. Since most Artificial Intelligence artefacts belong to the “literary, scientific or artistic domain” anyway, and are the result of at least some “human intellectual effort”, however remote, in practice the focus of the copyright analysis is on steps 3 and 4.

While Artificial Intelligence systems play a dominant role in the execution phase, the role of human authors at the conception stage often remains essential. Moreover, in many instances, human beings will also oversee the redaction stage. Depending on the facts of the case, this will allow human beings sufficient overall creative choice. Assuming these choices are expressed in the final Artificial Intelligence-assisted output, the output will then qualify as a copyright-protected work. By contrast, if an Artificial Intelligence system is programmed to automatically execute content without the output being conceived or redacted by a person exercising creative choices, there will be no “work”.

Authorship status will be accorded to the person or persons that have creatively contributed to the output. In most cases, this will be the user of the Artificial Intelligence system, not the Artificial Intelligence system developer, unless the developer and user collaborate on a specific AI production, in which case there will be co-authorship. If “off-the-shelf” Artificial Intelligence systems are used to create content, co-authorship claims by Artificial Intelligence developers will also be unlikely for merely commercial reasons, since Artificial Intelligence developers will normally not want to burden customers with downstream copyright claims. We, therefore, expect this issue to be clarified in the contractual terms of service of providers of such systems.

  1. Conclusion

In conclusion, the EU copyright framework is generally suitable and sufficiently flexible to deal with the current challenges posed by Artificial Intelligence-assisted creation. Producers of Artificial Intelligence-assisted output will in many cases enjoy copyright protection. Moreover, “authorless” Artificial Intelligence output might still qualify for protection against misappropriation under less demanding IP regimes, such as neighbouring rights and sui generis database protection,[9] or other doctrines such as trade secrets and unfair competition.[10] In this light, regulatory intervention to extend copyright protection beyond the current EU rules would be justified only if the solid empirical economic analysis were to reveal that the absence of protection harms overall economic welfare in the EU.
According to the data above, it can be understood that the legislation in the field of protection of intellectual property rights is developing every day and is spreading to more countries, which leads to the globalisation of this branch of law. Consequently, new methods of protecting the intellectual property rights of neuro-artists will appear. But this is possible only after a qualitative codification of the norms, which is currently on the path of its development.
“Security is a process, not a result.” Bruce Schneier – American cryptographer, writer and computer security specialist. Author of several books on security, cryptography and information security. Founder of cryptographic company Counterpane Internet Security, Inc., member of the board of directors of the International Association for Cryptologic Research and member of the advisory board of the Electronic Privacy Clearinghouse, also worked for Bell Labs and the US Department of Defense).

 

Citation

[1] ING. The next Rembrandt. URL: // https://www.nextrembrandt.com (text: electronic).

[2] Gervais (2019), p. 19.

[3] Senftleben and Buijtelaar (2020); Ana Ramalho (2017), p. 12.

[4] Kalyatin V.O. Copyright objects created using a computer // Patents and licences. 2011. No. 5. S. 22-25.

[5] Tsymbalova Y. Y. Copyright in works created by computer programs. / Collection of works of graduates of the Russian School of Private Law, dedicated to the 90th anniversary of the birth of Viktor Abramovich Dozortsev. Publisher: – Moscow: Statute, 2020. S. 615.

[6] Cohn G. Artificial Intelligence Art at Christie’s Sells for $432,500. URL: // https://www.nytimes.com/2018/10/25/arts/design/ai-art-sold-christies.html (accessed 03/30/2021).

[7] Butler T.L. Can a computer be an author – copyright aspects of artificial intelligence. / Hastings Communications and Entertainment Law Journal. No. 4, 1982. P. 744-745.

[8]  Compendium of U.S. Copyright Office Practices. Third edition, 2014. P. 22.

[9] Hugenholtz et al. (supra note 7)

[10] Hilty et al. (2020), p. 1; Scheuerer (2021)

 

The articles on the LAWELS platform are not, nor are they intended to be legal advice. You should consult a lawyer for individual advice or assessment regarding your own situation. The article only reflects the views of the author.