Violeta Arnaiz Medina. Director of Intellectual Property, AI and Software at PONS IP

Since the dawn of time, human beings have demonstrated an irrepressible creative drive that has transcended eras and styles. To this end, they have used tools that, in their constant evolution, have largely shaped the way we create—even the way we think.

But how do we now differentiate between art created by a human author and art generated by a tool that, for the first time, is capable of emulating this ability that has, until now, been uniquely ours? We already know that intellectual property legislation reserves copyright for human creators, and it is not possible for a machine author or an AI author to claim any legal monopoly over the outputs produced by these systems.

But it turns out that artists, always at the forefront, are determined to play with technology and even to hybridise with it. So what do we do with this “hybrid” art? Where does AI that generates outputs based on generic prompts end, and where does the creator with superpowers, the one who puts the tools at the service of their artistic cause, begin?  As doctrine has long indicated, what matters is who is in charge. Who directs the creative process and makes the decisions that condense the originality of the content created. The devil is in the details, and copyright, now more than ever, is too. 

In this sense, the very recent judgement of 13 February 2026 (142 C 9786/25) of the Local Court of Munich (Amtsgericht München)[1], which establishes a series of criteria to navigate these turbulent waters and help us determine who—the human or the AI—is at the helm of the creative process, is very interesting for its pioneering but also clarifying nature. The trigger is a lawsuit—somewhat prefabricated, according to the Court’s suspicion[2]—in which one party accuses another of using the following logos without permission:

As a starting point, the judgment recognises that “it is conceivable that copyright protection exists for the results of AI systems created as a result of human intervention”.In other words, the case of a human creator assisted by AI and benefiting from copyright is, in theory, possible. This intervention referred to in the ruling can be carried out through “sufficiently individualised prior adjustments in the programming of the process of creating the specific content” and/or through the iterative input of instructions or prompts into the system.

However, the results of this process will only be protectable “to the extent that, despite the software-controlled process, human creative influence continues to be exercised”. This influence, which must be expressed through “free and creative decisions, with the author adding their personal touch to the result”—as case law has unanimously established—will occur when “the use of the AI model is closer to an auxiliary tool than to an independent creation instrument”. And this, according to the ruling, will only happen when the creative elements incorporated into the prompt “dominate the result in such a way that the object as a whole can be considered an original creation of its author”.

Once this framework for action has been established, the ruling analyses the plaintiff’s actions one by one. And it concludes that, in these cases, “the creative design, that is, the exact selection of the elements incorporated into the illustration, has been left to the rules of AI”, which prevents the resulting works from being protected. We highlight some of the main ideas or criteria of this analysis:

• The act of entering prompts or instructions into an AI system will not generate copyright when these are general or open. Copyright does not protect ideas. Therefore, with regard to the content of the instructions, general indications such as the following (“Design an original and abstract logo”, “The design should be modern, minimalist and clearly original”, “Style: clean and flat design with customised geometric abstraction”, “Can you make the white leather hand more feminine?”, “Add a more realistic touch to the hands”) cannot encapsulate the creativity of the work that is generated as a result. Or, in the words of the court itself, “they do not allow any conclusions to be drawn about the type and appearance of the result, as they do not reflect any human creative decision of their own”.

• The number of prompts entered is not a sufficient indicator to determine that there is sufficient human contribution to the final result. For example, the fact that a complex prompt of more than 1,700 characters was entered for the creation of the illustration entitled “Envelope depicted in front of a building with columns” is irrelevant. This is because, as is well known, copyright does not reward or protect effort, time spent on creative activity, or investment of any kind (financial, effort, personnel, time, etc.).

• For the same reason, it is irrelevant whether a premium/paid or free version of the AI tool is used, because even though this could be indicative of an effort or interest in creating the specific content, that effort or interest is not protected by copyright.

• The iterative nature of the process, i.e., selecting proposals made by the AI system, even the action of retouching them through successive instructions, is not, in itself, sufficient to understand that we are dealing with a personal creative creation, if the indications are general or open and do not reflect human creative decisions of their own.

For all the above reasons, the judgment concludes that, having analysed the process of creating the three images, “the technical activity of the AI largely prevails over the human creative influence of the claimant”, which determines the full dismissal of the claim.

In conclusion, if it is the rules of AI—and not the human user—that are in control of the creative activity, it is not possible to apply copyrights over the generated content. Determining when this occurs and when it does not will always be a case-by-case basis, according to criteria that will be determined by case law and that this ruling helps us to begin to outline. This is without forgetting that what should be understood by work—which is directly linked to the requirement of originality—is an autonomous concept of EU law, so it will ultimately be the Court of Justice of the European Union that will be responsible for defining it in these contexts of hybrid art. Hard work.

[1] https://www.gesetze-bayern.de/Content/Document/Y-300-Z-BECKRS-B-2026-N-1513?hl=true

[2] “It is true that, in this constellation, there are indications that the parties are continuing the litigation (also) for scientific reasons,” states the ruling in section 16 (our translation), concluding that, nevertheless, a response to the controversy, which exists regardless, is necessary.