As it is already evident that we are living in a digital revolution, in which the scope of artificial intelligence (hereinafter: AI) is being bid daily, questions of authorship of the works that officially created the computer program have arisen.

In the following text, we will be based on the relationship of the AI ​​system, i.e. algorithms that process and analyze existing photos, texts, and music and use them to generate new ones and normative concept of work of authorship. Systems like ChatGPT, DALL·E, and Amper Music are globally popular and according to current information, there are about 100 million ChatGPT users who have more than 10 million queries daily. When these systems answer a query, with each answer they create certain works, which give rise to numerous questions.

Who owns the work?

The Law on Copyright and Related Rights (hereinafter: the Law) defines a work of authorship as an author’s original spiritual creation, expressed in a certain form, regardless of its artistic, scientific, or some other value, its purpose, size, contents and way of manifestation, as well as the permissibility of public communication of its content, and the author as a natural person who has created a work of authorship.

The key difference between a work of authorship and a work created “with the help”; of AI is that the Law defines the author exclusively as a natural person, while algorithms and programs are used to process data and generate new results based on predefined rules and parameters in which case, the created work is not the work of man, but is the result of the aforesaid technological AI process.

Also, according to Serbian Law, the work of authorship must be a “spiritual creation”, which is closely related to the fact that the author can only be a natural person – a man. The spiritual content of the work of authorship must be expressed in the work itself.

Is AI capable of creating original work?

The Law does not define originality, which is a prerequisite for the existence of a work of authorship and an author, but taking into account judicial practice and legal theory, one can come to the conclusion that a work is original if it was created independently, without copying someone else's work, an original work is an individual work. The works created by AI are such that they are mostly a combination and processing of existing data, with the application of algorithms. In this sense, these works are more the result of mathematical calculations than a creative process. Everything that seems “original” and “creative” to us is the collection, analysis, and processing of existing data. The technological development of AI has still not reached the originality inherent in human creativity, and whether this will happen in the future is the question left.

Copyright infringement by AI

The question is being increasingly raised around the world whether AI systems that generate photos and music can infringe copyright. In particular, these systems create photos or songs that mimic the author’s style, which may cause copyright infringement. However, the creators of AI systems claim that this process of collecting data to generate new works of art can be considered fair use and that this doctrine could be applied in appropriate cases.

Current approaches and the future of AI

Court practice regarding the potential authorship of AI is still developing, as this issue has emerged relatively recently and laws are adapting to new technologies. For the time being, the courts apply existing legal frameworks that are fairly harmonized regarding the definition of author and work of authorship.

AI generator Deep AI states on its terms of service page that all content created using its AI tools is free from copyright for any type of use, personal or commercial.

The US Copyright Office on 16.03.2023 adopted a policy confirming its previous statement – only humans can be authors, not AI. Furthermore, the author is not the natural person making the request, according to which the AI ​​later generates works, but artistic, musical, and other works created using AI may be eligible for copyright registration, under the condition that there is enough “human authorship”.

There are also different approaches, in countries such as Hong Kong (SAR), India, Ireland, New Zealand, and the UK, the authorship of the work created by artificial intelligence is assigned to the developer. This approach has been implemented by the Copyright, Designs and Patents Act of the UK, which stipulates that in the case of a literary, dramatic, musical or artistic work which is computer-generated, the author shall be taken to be the person by whom the arrangements necessary for the creation of the work are undertaken. This provision means that the developer who created the AI ​​that creates the works can be considered the author of those works. Although this solution is not applicable in all countries and all situations, it represents one of the approaches in solving the issue of authorship in the context of AI.

Also, the issue of copyright infringement by AI systems is not a simple one, as fair use is applied in different ways depending on the legal system. For example, while the concept of fair use applies in the US, it is unknown to the EU legal system. As such, the issue of copyright infringement by AI systems is very complex, and experts in the field note that it must be resolved in the courts.

Further research and development of the legal framework will likely be required to ensure copyright protection in an age when AI is becoming more and more present in the creation of art, and other creative works with the question of whether the works are works of authorship.

Ultimately, the balance between protecting copyright and enabling innovation in AI will be key to the viability of this field in the future, which means it should be ensured that copyright does not limit progress in AI while also protecting authors and their rights.