By: Lorena Pinto Sanabria
In recent years, artificial intelligence (AI) has ceased to be a futuristic concept and has become a creative tool used in everyday life. From images and music to texts, videos, and software, AI-assisted works have gained increasing relevance in the creative field. However, this progress raises a fundamental question in the field of copyright law: Who is the legitimate author of a work created with artificial intelligence?
The current legal framework, both in civil law systems and common law jurisdictions, is based on a traditional notion of authorship based in human creativity. For instance, the Mexican Federal Copyright Law defines the author as the natural person who creates a work. This definition does not contemplate the participation of machines or automated systems in the creative process, only human beings.
In light of this, three main theories have emerged regarding how authorship should be addressed in the context of artificial intelligence:
1. The creator of the AI algorithm as the author
Some legal scholars and engineers argue that the author of an AI-generated work should be the individual who developed the algorithm or artificial intelligence model. Their reasoning is that the programmer designs the logical and functional structure that makes content generation possible, and thus their creative and intellectual contribution is decisive in the creation of the work. This view aligns with the idea that the software is an extension of human creativity.
However, this position has limitations, especially when dealing with general-purpose generative AI systems that were not specifically designed to produce particular works. In such cases, the link between the programmer and the resulting work may be too remote to justify authorship.
2. The user of the AI as the author
Another theory suggests that the user of the AI should be considered the author, as they define the purpose, input the prompts, select the outputs, and often edit or adapt the generated content. Under this perspective, AI is regarded as a tool, comparable to a camera for a photographer or a paintbrush for a painter, rather than an autonomous creative entity.
This position currently enjoys the widest acceptance in practice, particularly in countries that recognize the importance of substantial human intervention in the creation of a work as a prerequisite for copyright protection. However, challenges remain in determining the degree of human involvement necessary for a work to be deemed protectable.
3. The AI itself as the author
A third, more controversial, theory suggests that artificial intelligence could be recognized as the author of its own creations. This idea has gained traction as AI systems are now capable of generating works that appear comparable to those created by humans. Nonetheless, this view faces fundamental legal, ethical, and philosophical obstacles, as machines lack legal personality, consciousness, and volition, traits traditionally associated with the human notion of authorship.
Additionally, a major concern relates to the training of AI systems. In many cases, it is unclear what data or content was used to train the models, which may involve the unauthorized incorporation of previously copyrighted material. This lack of transparency not only complicates any claim of legitimate authorship by the AI itself, but also raises serious legal risks for both users and developers who rely on such outputs.
Recognizing AI as an author would require a significant overhaul of current legal frameworks and would also trigger complex debates about the ownership of economic rights, liability for infringement, and the possible commercial exploitation of works generated without human involvement.
Cases in Mexico
Specifically in Mexico, in February 2022, an individual filed an application before the National Copyright Institute to register an image generated by artificial intelligence, designating the AI as the author of the work. The Authority denied the registration, stating that Mexican legislation only recognizes natural persons as authors.
The applicant challenged the decision, and the Court of Appeals upheld the denial. The court concluded that, under the Mexican Copyright Law, an AI system can not be considered the author of a work, as authorship is limited to human individuals.
This case constitutes the first judicial resolution in Mexico to explicitly address the inadmissibility of registering works where AI is named as the author.
As a result, the prevailing legal standard in Mexico is that AI cannot be recognized as an author or co-author, and that creative works generated without human intervention are not eligible for copyright protection. This imposes a limitation on creators and users of these technologies, but it also opens the door to potential legislative reforms that might recognize the importance of regulating substantial human involvement in AI-assisted creative processes.
In conclusion, the debate over authorship in AI-generated works remains open and, for now, lacks a definitive answer. While technology advances rapidly, the law has yet to offer analogous solutions adapted to this new reality. What is clear, however, is that this evolving landscape presents an opportunity for legal, ethical, and philosophical reflection, requiring us to rethink our fundamental concepts of creativity, intellectual property, and the relationship between humans and machines.
Contact
Mauricio Chozas
Lorena Pinto
