A work generated by AI and copyright – new case law

Publication date: December 20, 2023

With the development of technology, artificial intelligence began to impact almost all areas of our lives, starting from performing tedious simple tasks, it has developed to the point that it is able to create art, including graphics and music. With this progress, new issues and problems arise related to this AI activity. One of the main current problems is the question of who holds the copyright to works created in this way and whether such works are covered by copyright protection at all. Questions arise such as: can artificial intelligence be considered a creator? Do works created by artificial intelligence have copyright? What human contribution is sufficient for copyright protection of a work? Who is the creator of the work and who should have the copyright?

US Court ruling regarding works generated by artificial intelligence

            The discussion regarding copyright protection for works created by artificial intelligence recently reignited again when the Federal District Court for the District of Columbia ruled in Thaler v. Perlmutter. In this case, the plaintiff filed an application for registration of a work created autonomously by a machine and claimed copyright in the work as the owner and constructor of the machine. The US Copyright Office, which considered the application, issued a negative decision, and further attempts to register the work led the parties to court proceedings. The main question raised during the case was – Is a work independently generated by an AI system subject to copyright? In response to this question, the court clearly stated in the ruling that such a work is not subject to copyright protection. However, the Court did not consider during the hearing new facts revealed by the plaintiff (they provided information about the creative process and human contribution to the creation of the work) because the judicial review was limited only to the same information that the Office had at its disposal – in this case it was the plaintiff’s claim to copyright of a work that was “created autonomously by a machine” and based his claims on the fact that he was its owner (constructor). However, the Court noted that in the future the criteria will be assessed as to how much human input is necessary to qualify a user of an AI system as an author. In the case in question, the complainant cited the AI tool as the author and did not prove his contribution to the creation of the image. Currently, a work created by AI is not subject to copyright protection, however, as most lawyers predict, in the future there will be new legislative guidelines clearly specifying to applicants what the limitations of submitted applications are, and case law and legislation will continue to evolve – therefore the current legal status of the works created by artificial intelligence in the US jurisdiction makes it clear that such works are not subject to copyright protection.

European and Polish case law regarding works created by AI

            The question arises how European case law approaches this matter and what view is presented by Polish courts in the context of the Copyright and Related Rights Act of February 4, 1994. First of all, it is worth discussing the view presented by the Court of Justice of the European Union, which in the case of Eva-Maria Painer v. Standard VerlagsGmbH et al. of December 1, 2011 clearly indicated what the subject of copyright protection is, namely, it is the result of the author’s individual creativity, provided that it reflects the creator’s personality and the free and creative choices made during its implementation. However, in the judgment of September 12, 2019 in the case of Cofemel-Sociedade de Vestuario SA v. G-Star Raw CV, the court supplemented the above-described position by stating: “Where the execution of the object is conditioned by technical considerations, rules or restrictions that leave no room for creative freedom, the subject matter cannot be regarded as exhibiting the originality necessary to constitute a work.” The position of the EU copyright system around the anthropocentric view and the personal nature of the work is additionally strengthened by the opinion of the then Advocate General, V. Trstenjak, who, issuing her opinion in the case of Painer v. Standard VerlagsGmbH, stated unequivocally that only human works are subject to protection. Such a firm view, regulated in the EU in the spirit of current changes, does not completely eliminate the possibility of copyright protection of works created using AI. The literature indicates that when a work contains a feature of authorial work that can be directly attributed to the author of the program or another person, such works can be protected. The European Parliament also commented on this issue in its resolution of October 20, 2020 on intellectual property rights in the field of artificial intelligence technology development (2020/2015[INI]), clearly pointing out that human creativity supported by AI should be distinguished from creativity generated only by AI. It argued that in the first case, the current intellectual property framework applies, while in the second case, it is necessary to implement regulations in this area. Moreover, the EP indicates that regulations on AI should take the form of a regulation in order to standardize the rules throughout the EU. It is worth mentioning here the quite controversial views that are currently put forward in the doctrine. According to them, due to artificial intelligence having cognitive abilities, it may also have creative ability, i.e. be a creator within the meaning of copyright law, regardless of whether it will have broader or more general legal capacity and subjectivity[1]. However, under the current provisions of copyright law, it is not possible to provide copyright protection to an object created using AI, because the basic factor distinguishing computer-generated works from works created using a computer program is the lack of a specific person’s creative contribution to the creation of the work – all features of the work that determine the granting of copyright protection if they were created by a human being, are the result of the work of a machine[2]. Given the current concept of a work and its authorship, neither the user of a given program nor its creator can be considered the authors of a computer-generated object. A reflection of this view can be found in the jurisprudence of Polish courts, including the Judgment of the Court of Appeal in Łódź (case files number: I AGa 54/18), which states: A work is an intangible legal good that is, first of all, the result of creative, i.e. original activity of intellectual property of a person (the creator of the work) or a team of people (co-creators of the work) and, secondly, it also has the mark of individuality, i.e. it is characterized by features that allow it to be distinguished from other, previously created intangible goods of this type. Pursuant to Article 1 of the Copyright Act of 1994, the first condition for obtaining protection under copyright law is the achievement of the effect of creative activity, i.e. the external manifestation of a specific human action[3]. As P. Ślęzak points out in his commentary on the act, the doctrine agrees that creative work can only be performed by humans, therefore other effects or manifestations of behavior, including animals, nature or devices, remain outside the scope of protection. In the light of such regulations, current works created by AI are not protected, therefore there are no restrictions on their exploitation by other people. Even though the Act itself, in Article 1(1), does not explicitly stipulate that the title of creator is reserved only for natural persons, the case law and literature in this area clearly state that “the original source of copyright is human creative activity”[4]. It is pointed out that only a person is able to give such features to a work that it meets the criteria for protection; if such criteria are met, it is necessary to establish the existence of a connection between the activity of the person claiming the status of creator and the creation of the subject of protection.[5] Therefore, when examining whether a given work meets the conditions for protection, it is necessary to always examine how the work was created and what human contribution was to the creation of the work. To conclude, the current trend puts pressure on legislators to take action to create regulations regarding such works that clearly define the minimum human contribution to the creation of the work, so that it is covered by protection, and to create objective, measurable and verifiable criteria of creativity allowing to determine this contribution – such aspirations can be found, among others, in the actions of the European Parliament, which demonstrates that “technological creativity created by AI technologies must be protected by intellectual property rights in order to encourage investment in this form of creativity and increase legal certainty for citizens, businesses and inventors who are currently among the most frequent users of AI technology[6]. If this state of affairs is left unchecked, it may lead to problems in this respect, works created by AI belonging to the public domain open up opportunities for abuse – “it creates an incentive to conceal the fact of using artificial intelligence in the creation process and to attribute authorship of these works by natural persons”[7].
As a result of the search for a solution to reconcile the development of AI and copyright, including stopping the market marginalization of the position of human creativity, three groups are emerging proposing specific solutions. The first group includes concepts according to which computer-generated works, i.e. using AI, should be protected by copyright. However, to make this possible, it would be necessary to expand the circle of entities capable of obtaining the status of an author within the meaning of copyright law and to regulate computer-generated works in the context of who would be the subject of copyright for this type of works. However, such a solution in Poland is controversial – “depersonalization of the premises determining the creation of a work means a departure from the assumptions of copyright protection of the work, understood as a manifestation of activity of an individual, personal nature, which in turn implies a re-evaluation in the system of norms that are the basis of European legal culture.”[8] The question arises whether such action aimed at extending copyright protection to works generated by AI is justified. Therefore, the second group of concepts presents a view that opposes copyright protection for works created by artificial intelligence, but at the same time notes that compensation should be provided to entities involved in the development of artificial intelligence (AI) systems for the financial and organizational expenses incurred. Therefore, they propose the introduction of regulations of a similar nature, and there are suggestions in the literature regarding the creation of a new category of related rights or the establishment of a special exclusive right (sui generis) for this type of works. The third group includes concepts according to which works generated by AI should remain in the public domain. What connects these concepts is the belief that failure to transfer such works to the public domain does not involve the risk of inhibiting progress in the field of artificial intelligence, at least not a risk that would justify legislative intervention. It is argued that if an artificial intelligence system can be protected by copyright as a computer program, then the financial and intellectual contribution of copyright holders is already sufficiently rewarded. There are no obstacles, for example, for entities holding copyrights to a generative AI system to derive financial benefits by granting paid licenses for its use.

            To sum up, the debate around copyright protection for works generated by artificial intelligence raises questions about the identity of the creator, the nature of protection and the impact on the development of this technology. It is necessary to develop clear regulations to respond to new challenges while maintaining a balance between technological progress and the protection of creators’ rights.

[1] P. Księżak, S. Wojtczak, Copyright towards artificial intelligence (an attempt at an alternative view), PIP 2021, No. 2, pp. 18-33.

[2] P. Juściński, Copyright in the face of the development of artificial intelligence, ZNUJ. PPWI 2019, no. 1, pp.5-44.

[3] P. Ślęzak, Act on copyright and related rights. Commentary, Warsaw, 2017.

[4] Judgment of the Supreme Court of February 14, 2014, II CSK 281/13.

[5] A. Bar, Copyright in the era of artificial intelligence. Remarks on the history of “Portrait of Edmond de Belama”, 2022.

[6] European Parliament resolution of 20/10/2020, 2020/2014(INL), point 15.

[7] A. Bar, Copyright in the era of artificial intelligence. Remarks on the history of “Portrait of Edmond de Belama”, 2022.

[8]I. Oleksiuk, Axiological assumptions of copyright protection of creativity in the light of the development of artificial intelligence, Acta Iuris Stetinensis, 18 (2), 2017, pp. 245-262.