Legal regulation of intellectual property for objects created with the use of AI

Legislation in the field of IP rights protection

Legislation in the field of intellectual property rights protection for works generated by artificial intelligence is new and is in the process of active development.

According to Art. 435 of the Civil Code and Art. 5 of the Law on Copyright and Related Rights, AI is not a subject of copyright. That is, it cannot be considered an author and is outside the scope of legal regulation.

An author can be an individual who has created a work through his or her creative activity. Personal non-property rights do not arise for non-original objects generated by a computer program. Such works are considered non-original objects generated by a computer program and are protected by a special kind of law (sui generis).

Nevertheless, in 2024, the Ukrainian National Office of Intellectual Property and Innovation (IP Office) registered intellectual property rights to the collection “Poems of an unfinished war” and the children’s book “Rufus’s enchanted adventure”, which include images created by artificial intelligence, and the right holders of the registered objects are individuals, the authors of the relevant texts. This does not mean that such AI objects receive legal protection, but rather a disclaimer regarding the inclusion of such works in the collection. When registering copyright, the application form must specify the objects used in the new work (e.g., a part of the code that is open source or licensed, objects of other authors, etc., and a warning about the legality of using such objects).

EU cases

In the European Union, two patent applications were filed for registration in 2018: for a food container and devices and methods for attracting increased attention, which were rejected by the European Patent Office (EPO). The reason is that according to the European Patent Convention (EPC), only a person, not a machine, can be an inventor. In the applications, the artificial intelligence “DABUS” was listed as the inventor. The applicant claimed that as the owner of the machine, he had received all rights to its development. However, in its conclusions, the EPO emphasized that only an individual can be an inventor, which is a generally accepted international standard. In addition, the legal status of an inventor has legal consequences that machines cannot have.

The EU’s final report on AI trends and developments in 2020 states that works are subject to protection if they are “original in the sense of being the author’s own intellectual creation”. In their opinion in the Painer case, Advocate General Trstenjak and the Court of Justice concluded from the content of Article 6 of Directive 93/98 and Directive 2006/116 on the terms of protection that “only human works are protected, which may also include those for which a person uses a technical means, such as a camera”. This requirement for human effort excludes from the possibility of legal protection results that were created without any human intervention.

However, the European Union adopted the AI Act, which came into force on August 01, 2024. Article 107 of this Act provides that providers of generative artificial intelligence models will have to provide a detailed summary of the content used for training in a comprehensive manner, which will allow authors or parties with legitimate interests to exercise and enforce their rights under EU law. This could also help to further recognize works created by artificial intelligence while ensuring that the copyrights of those whose works have been used to train it are respected.

USA

Copyright law in the United States generally requires an individual to be the author, a principle that stems from the Copyright Act.

On the other hand, in more recent cases involving co-authorship, the courts have examined the required level of contribution to qualify as an author.

For example, the US Supreme Court considered this issue in the case of Community for Creative Non-Violence v. Reid (CCNV), where it was decided who is the author of a sculpture: the non-profit organization that conceived it or the artist who created it. The court ruled that the artist was the author, since he had made the sketches and embodied his creative idea in a material form. In its decision to remand the case, the D.C. Circuit noted that commissioning a sculpture and providing detailed instructions is not enough to recognize authorship, as such contributions are not covered by the relevant protection. Although the result created by artificial intelligence cannot be considered a joint work of the user and the AI system, joint authorship provides a useful analogy for assessing whether a party has made sufficient creative contribution to be recognized as an author.

Types of human contribution to AI-generated results

In addition, when using AI, a person usually applies instructions that let the AI understand what is expected of it (prompts) and if only such detailed instructions are provided without control over their implementation, it is not enough to recognize authorship, i.e. co-authorship in this way with AI demonstrates that the user does not control how his or her idea is transformed into an expressed work. As the court noted in its decision, a person who merely describes to the author how the work should be commissioned or how it should look is not a co-author. Therefore, to be a co-author, one must make a contribution that falls under copyright.

In the creative process, there are three types of human contributions to AI-generated results:

• Prompts that guide the AI system to create a result.
• Explicit (creative) inputs that can be recognized in the results created by AI.
• Modifications or reordering of AI-generated outputs.

Conclusion

There is, therefore, an important distinction between using AI as a tool to assist in the creation of works and using AI as a replacement for human creativity. In cases where AI helps to expand human expression, this does not limit copyright protection.

Although the law does not yet contain clear provisions on copyright for objects created with the help of artificial intelligence, current discussions and legal trends indicate that they may be recognized in case of significant human creative contribution. If AI played only an auxiliary role in the creative process, and the result contains original elements, the user of such a tool can potentially claim authorship. At the same time, the limits of this contribution remain the subject of active discussion.

Published by UNBA

Link to the article