Neural Network and Copyright: The Limits of Technical Assistance and Creative Contribution

In 2025, many media reported that the State Duma of the Russian Federation wants to secure copyright for works created by artificial intelligence. In particular, it is planned to amend Article 1259 of the Civil Code of the Russian Federation and secure copyright for works generated using a neural network. There is also an increase in litigation related to copyright, where the subject of litigation is content created using generative models.

Thus, the need for legislative regulation of the issue is due not only to the need to protect the rights of the authors of the relevant content, but also the possibility of their protection, in particular, in disputes over the rights to content with owners of technologies and copyright holders of source data, in disputes over the illegal use of content by third parties, etc.

Given the high relevance of the topic and the debatable nature of the issue, let's try to understand under what conditions the result of intellectual activity created with the help of artificial intelligence can be an object of intellectual property rights.

When work is recognized as creative, and the result is intellectual

According to the current rules, the author of the result of intellectual activity and, accordingly, the person who initially has the exclusive right to this result, can only be an individual - a citizen whose creative work created such a result (Clauses 1 and 3 of Article 1228 of the Civil Code of the Russian Federation).

On the contrary, the Civil Code of the Russian Federation and the explanatory acts of the Supreme Court of the Russian Federation proceed from the opposite. Firstly, the legal protection of copyright objects is based on the principle referred to in the doctrine as "aesthetic neutrality". The relevant presumption is enshrined in the clarifications of the Supreme Court of the Russian Federation, according to which the results of intellectual activity are presumed to be created by creative work, until proven otherwise, and such criteria as novelty, uniqueness and (or) originality do not affect the recognition of the results as creative.

Secondly, the law expressly specifies cases when participation in the creation of the result of intellectual activity is not recognized as a creative contribution. In particular, they include technical, consulting, organizational, material assistance or assistance, as well as control over the creation of the result (Clause 1 of Article 1228 of the Civil Code of the Russian Federation). As a rule, most of these cases of participation do not cause controversy, since they are quite easy to separate from the actual creative contribution to the creation of the result of intellectual activity.

Disputes arise in terms of technical assistance - the qualification of specific actions as technical or creative. In this regard, the Supreme Court of the Russian Federation explained that technical assistance is understood as the performance of auxiliary operations that are not of a creative nature, for example, the selection of materials, the drawing of schemes, diagrams, graphs, the production of drawings, photographs, models and samples, the performance of calculations, the preparation of documentation, the conduct of experimental testing, etc. In addition, it is noted that the creative nature of the creation of a work does not depend on whether the work was created by the author himself or with the use of technical means, but the results created with the help of technical means in the absence of the creative nature of human activity are not objects of copyright.

However, in a situation where the results of intellectual activity are generated with the help of artificial intelligence, these clarifications of the Supreme Court of the Russian Federation are clearly insufficient. The fact is that in this process, a neural network is not just a technical tool that can be used, for example, for the automatic selection of structured information (materials), but also performs a wider range of tasks that, by the nature of their functions, can be conditionally qualified as "co-authorship".

For example, from the definition of artificial intelligence, of which neural networks are a variety, it follows that the technology of AI is based on the imitation of human cognitive functions (including the search for solutions without a predetermined algorithm), and the results of artificial intelligence are comparable to or superior to the results of human intellectual activity.

Since in the past few years there has been a massive use of neural networks to create creative content, which is formally indistinguishable from works created by humans, there are discussions about whether artificial intelligence can be considered the author of works and whether works created with the help of neural networks should be legally protected.

In our opinion, artificial intelligence (including a neural network) can be neither the author nor the co-author of the result of intellectual activity, nor any other subject of law. However, we believe that the question of what is the creative contribution of the person who creates the work with the help of a neural network in this case. We believe that the answer to the question of whether the result of intellectual activity created with the help of artificial intelligence can be an object of intellectual property rights directly depends on the solution of this issue.

Let us try to look at the problem from the point of view of the approaches formed in doctrine and judicial practice.

How creativity is evaluated when using technical means

Researchers note that for copyright there is no need to specifically formulate a legal definition of creativity, taking into account the different views of lawyers, philosophers, psychologists, etc. It is more correct to define the meaning of this term based on its usual meaning in everyday language, in which creativity is understood rather as a form of human activity, which is expressed in various types of activity and which is engaged not only by masters in the field of literature and art, but also ordinary people.

Thus, it is obvious that the creation of material results with the help of technical means operating in a fully automatic mode cannot lead to the creation of an object of intellectual activity. This is confirmed by the position of the Supreme Court of the Russian Federation, which, as an example of the lack of human creative activity, points to the results of photo and video filming operating in automatic mode of a video surveillance camera used to record administrative offenses.

In this case, the question arises about the correlation between the technical and creative parts, namely, how active a person's participation should be in order for the result to be recognized as creative and how such activity is expressed.

It seems that the basic approach to solving the issue can be found in the Recommendations of the Scientific Advisory Council at the Intellectual Property Rights Court, which were given on issues related to the definition of criteria for creative activity on the example of photography. It follows from the Recommendations that in order to resolve the issue of legal protection of an object, it is necessary to establish whether the author showed creative abilities through free and creative choice when creating it.

Thus, in relation to photography, the author's creative choice when creating a picture may consist, in particular, in the selection of composition, the choice of angle, subject, light, filters, and image processing.

With regard to an audiovisual work, even created with the help of AI, for example, deepfake technology, the creative choice of the author may consist in creating a script for the work, filming source materials, audio accompaniment, etc.

However, we believe that, following the basic approach, any work created with the help of any AI technology can be recognized as an object of copyright, if its creation is dictated by the free and creative choice of a person, even in a situation where the creative contribution of the latter seems insignificant relative to the contribution of AI.

Prompt as a free and creative choice

As a rule, when creating works using neural networks, the minimal, but basic unit of human participation is the prompt.

The word prompt comes from English (prompt) and in the IT field means a pre-formulated request or instruction for a neural network. Simply put, prompt is a task that a user gives to an AI model when describing a task.

It is interesting to note that according to the Gramota.ru portal, in 2024 the word "promt" became the word of the year in the IT category. At the same time, the evaluation criteria were not only the steady growth of its mention, but also the relevance of the phenomenon among specialized specialists, its importance for the industry, as well as the potential for development in the coming years.

There are various types of prompts for neural networks, including those that generate the creation of works. For example, textual (a request to write a literary work according to certain criteria, etc.), visual (commands for generating images, including a description of a scene, landscape or object), auditory (instructions for creating or processing sounds and music, including a description of the desired genre, mood, certain instruments, etc.).

The key factor is that in all cases the creation of such works is based on the free and creative choice of a person who formulates the appropriate prompt – a specific task or set of instructions for artificial intelligence, which technically translates the ideas, thoughts and images of the author of the prompt into material form.

Thus, we believe that neither in law, nor in practice, nor in the logic of the relevant relations there are any obstacles for the result of intellectual activity created with the help of artificial intelligence (neural network) to be recognized as an object of intellectual property rights, and the author of the trademark to be granted legal protection.