Antonio Tejeda Encinas, president Comite Euro Americano Digital Law
The dizzying growth and popularity of generative artificial intelligence, capable of creating text, images, and videos from natural language instructions (prompts), have sparked a debate on the nature and ownership of this content: Should it be considered works protected by Intellectual Property? And, if so, could artificial intelligence be considered the author?
The straightforward answer is NO, given that, according to the Spanish Intellectual Property Law (LPI), the author is the natural person who creates a literary, artistic, or scientific work. Although in certain situations legal persons may benefit from this protection, an AI has no legal personality and, therefore, cannot receive this recognition.
Beyond this basic definition, both the LPI and case law have established a legal framework defining the concepts of author and work. We have considered whether, within this framework, an artificial intelligence could be regarded as an author. The LPI describes a “work” as «original literary, artistic, or scientific creations expressed by any medium or format, tangible or intangible, known or yet to be invented». The Supreme Court requires a «minimum of originality and creative height» for something to be protected by intellectual property. The Court of Justice of the European Union has interpreted “originality” as the «expression of the author’s personality» and “work” as «the author’s own intellectual creation, an expression of free and creative choices».
From this perspective, it is evident that content generated by an AI does not reflect free and creative choices, since it cannot create without a prompt from the user. In other words, the AI has no initiative or autonomy and only produces results after receiving a specific instruction. This is similar to commissioned works, such as a painting requested from a painter. However, in that case, the painter, although following instructions, imprints their personality on the painting; for example, a still life by Sorolla is clearly distinct from one by Picasso.
Therefore, the results generated by AI lack originality, since they depend on prior input. This means they cannot create something truly new. Although this could be compared to the influences any author has, the difference lies in the fact that human authors express their personality in their works, distinguishing them from those that have influenced them. Consequently, the second reason why AI-generated content lacks originality is that an AI cannot express personality, given that it does not possess this characteristic.
Having clarified that, under the current legal framework, an AI cannot be an author, the question arises as to whether it should be. Among the legal implications of this is that AIs would hold moral rights over their works. These rights include the power to decide on the disclosure of the work, to prevent any distortion or alteration that harms their legitimate interests or damages their reputation, and to withdraw the work from the market due to a change in their intellectual or moral convictions.
This would mean that the AI should be able to decide whether the generated content can be published, which is impossible, since it lacks free will: it cannot refuse to produce a result in response to a user’s command, unless prohibited by the developers who programmed it. Once the content is generated, the AI cannot decide on its use. Some AIs include in their terms and conditions an assignment of exploitation rights depending on what the user pays, but again, the decision rests with the company behind the AI and not with the AI itself. It could be understood that the authorization to disclose is implicit in the “acceptance” of the commission and, therefore, that the moral right is being exercised. However, there is no logic covering the right to withdraw the work in the event of a change in intellectual or moral convictions, since an AI lacks these intrinsically human qualities.
Therefore, if the legislator decided to grant some degree of protection as Intellectual Property to content generated by AIs, it should be considered as “performances” under the LPI. These are protected by related rights and do not confer moral rights. The nature of AI-generated content is more akin to mere photographs (those which, due to insufficient originality, do not have the character of photographic works protectable by copyright), in which the photographer is not considered an author, but a maker. This would avoid the problem of moral rights described above, while guaranteeing the exclusive rights of reproduction, distribution, and communication to the public.
Another position in the debate suggests that the user who writes the prompt should be considered the author. This discussion is reminiscent of the one that arose with the invention of photography, when the jurists of the time considered that a photo could not be a work protectable by copyright because it was made by a machine and the person behind the camera merely pressed a button.
We do not share this reasoning, since a photographer not only makes free and creative choices about what to capture and how to do so, but also controls the result. By contrast, a user of an AI, even if they make free and creative choices when writing the prompt, does not control the result generated by the AI; different AIs will produce different content in response to the same prompt.
For this reason, we consider that, in any event, if the user of the AI were the author of anything, it would be of the prompt. Thus, prompts would be regarded as literary works, just like computer programs. According to our LPI, a computer program is «any sequence of instructions or indications intended to be used, directly or indirectly, in a computer system to perform a function or a task or to obtain a specific result, whatever its form of expression and fixation»; a definition that closely resembles that of a prompt.
Under the current legal framework, neither can an AI be considered an author, nor can the content it generates be considered a work; moreover, the consequences of doing so do not fit within our legal system. The figures of performance and related rights are more suitable for protecting this content than those of work and copyright, respectively. However, we leave open the possibility that prompts may be considered literary works (if they are sufficiently original) and that the users who write them may be considered authors.
