Court Ruling on AI-Generated Works and Copyright Has Broad Impact

The Second Chamber of the Supreme Court of Justice of the Nation (SCJN) has unanimously ruled that works autonomously generated by artificial intelligence are not eligible for copyright protection under Mexican law.

Merits of the Case

The case originated in 2024, when a citizen submitted a request to the National Copyright Institute to register a digital work titled Virtual Avatar, created using the AI platform Leonardo. The request was denied based on the Federal Copyright Law (LFDA), which stipulates that only human creations can be recognized as protected works, as they must reflect originality and embody the author’s individuality and personality.

Following the rejection, the applicant challenged the decision before the Specialized Intellectual Property Chamber of the Federal Court of Administrative Justice (TFJA), arguing that excluding AI-generated works restricts creativity to the human domain and violates principles of equality, non-discrimination, and legal certainty. The challenge also cited international treaties to which Mexico is a party, including the United States-Mexico-Canada Agreement (USMCA) and the Berne Convention.

The TFJA upheld the rejection, prompting the applicant to file an amparo lawsuit before the Federal Judiciary. The Second Chamber of the SCJN took up the case and ultimately denied the amparo, reaffirming that works created solely by AI do not meet the originality requirements established under Mexican law.

Key Findings of the SCJN Decision

Copyright is a human right exclusive to natural persons, derived from their creativity, intellect, emotions, and lived experiences. Works produced solely by AI systems lack personal expression, a fundamental requirement for legal protection. Neither the Berne Convention nor the United States-Mexico-Canada Agreement recognizes AI as a subject of authorship. Thus, USMCA does not oblige Mexico to recognize copyrights to non-human entities or to extend the concept of authorship beyond what is established in the LFDA. This dismisses the idea that international trade or intellectual property commitments impose recognition of AI as a subject of moral rights. Articles 12 and 18 of the LFDA are constitutional, as limiting authorship to natural persons is considered an objective, reasonable, and internationally compatible standard. The principle of legal equality does not apply in this context, as humans and artificial systems possess fundamentally incompatible characteristics. The territoriality principle in intellectual property law reinforces that Mexican legislation governs registration requests submitted within the country.

Despite the warning about the inadequacy of the current legal framework, the SCJN held that Articles 3 and 12 of the LFDA are clear in establishing that only natural persons can be considered authors. "The author must be a natural person. It cannot be a synthetic or artificial entity," the body stated.

The ruling confirms that only human beings can be considered authors and only works created by humans are eligible for copyright protection. This position is consistent with prior decisions, including a 2022 ruling by the SCJN’s First Chamber, which held that legal entities cannot be authors, as creative action is exclusive to individuals.

Before this decision was ruled, the original draft declared that "the product of AI is an extension of humanity's creative capacity. Therefore, it is not subject to registration and, consequently, such products enter the public domain for free distribution." This conclusion was criticized based on the fact that the merits of the case were not focused on the effects of protection, but instead only on the authorship of the works generated by/with AI. The justices of the SCJN no longer mentioned this argument in the final ruling.

Nevertheless, the SCJN emphasized that “the creativity, originality, and individuality required by law can only arise from human experience, emotions, and intellect. Consequently, no artificial intelligence system, no matter how advanced, can meet these requirements.”

International Context

Mexico is not alone in this approach. In the United States, for example, copyright protection requires a human creative component. Applicants must disclose which parts of work were generated by AI. Works that modify AI-generated material or combine it with substantial human input may be eligible for protection. Examples of works that could be protected include those where AI-created material is modified or combined with materials from a human author who provides them for that purpose.

Legal and Technological Challenges Ahead

Faced with this reality, governments around the world, including Mexico, face a crucial crossroads: advance technological adoption to remain competitive, or attempt to protect jobs through regulations that could hinder economic and social progress. In this context, Mexico seems to be leaning toward the latter, which could ultimately condemn the country to falling behind in an era that demands immediate innovation and adaptation.

While this decision aligns with prevailing international standards, it leaves open important questions about the legal treatment of works created with AI assistance. The central issue is the degree and quality of human contribution required for a work to be protected. What happens when human input is significant? How much is enough?

This precedent does not define the scope of protection or permissible use for AI-assisted content, and a clear, definitive line may be impossible to draw given the complexity and variety of possible scenarios.

For artists, agencies, and technology companies, the central message is twofold. On the one hand, it confirms that artificial intelligence is not an author and cannot hold moral rights; on the other, it opens the door to registering hybrid works, in which a natural person directs, selects, edits, or transforms the algorithm-generated result to give it originality and a personal touch.