Artificial intelligence as a gravedigger of copyright?

Anyone who creates a text, image or piece of music in Germany is considered the author and has the corresponding rights. But what if not a human but an AI software created the content? Or to put it somewhat polemically: Is artificial intelligence the gravedigger of copyright?

First of all: In Germany there are currently no specific standards that deal with the use of “AI generators”.

However, the question arises as to whether the current standards and regulations are sufficient to regulate the situations created by AI systems. The central question is: Who is the author of the works of artificial intelligence?

Artificial intelligence as originator

Because the problem is that an AI can create works without a human artist. According to § 7 UrhG However, according to German law, the author must always be a natural person, because it says there that the author of a work is the “creator”.

As a prerequisite for legal protection, copyright law stipulates that a work is a “personal intellectual creation” and was therefore created by a natural person. This has two consequences:

In contrast to human authors, AI systems are not able to carry out an act of creation in the sense of copyright. AI systems are algorithmic machine entities that create works based on data and pattern recognition.

The question of whether a work created by AI is of sufficient level of creation and originality to be copyrighted is also problematic. Often, AI-generated works are based on existing copyrighted works that serve as training data. This can lead to a collision with existing intellectual property rights.

Taking these characteristics into account, the application of German copyright law to AI systems as authors is probably not possible. When asked about the author, one could think of the operator or developer of the program or the user who provides the input for the work to be created.

User Authorship

The assumption of authorship for the user would then be conceivable “if the influence of a person is so decisive that the result of the design process can still be attributed to him”. Now it is safe to say that a user can steer the result in the desired direction by adding various criteria or by reformulating the generated text. However, it is difficult to determine or measure the user’s own creative work in contrast to that of the AI.

However, it can still be considered whether the input command (“prompt” in English) falls within the scope of copyright protection. But then it would have to be clear which requirements must be placed on the entered character string. However, the combination of words in the input field of an image generator hardly reveals an individual level of creativity.

Authorship for the operator/developer

The operator or developer of the program would then still be someone who could be regarded as the author. This appears to be a conceivable approach, especially when one considers that the developer in particular enables the design of works through the program routine. However, if the AI ​​software “learns” on its own, it becomes clear that the contribution of the developer no longer plays a major role.

In addition, the developer would then be the author of possibly millions of texts, images or pieces of music without him knowing about it. Now active knowledge is not necessarily a decisive criterion, but it still seems strange when the developer of the software cannot even foresee how the software arrives at its results. However, attributing these results to the developer seems questionable.

Author: Artificial intelligence creates works and thus problems

These thoughts show the problems that arise from works created by AI. The problem becomes even greater when one considers how someone now wants to assert intellectual property rights based on AI-created works, especially when an AI-created work is further developed by a person. What solutions could we now discuss?

One could think of adapting copyright to the new realities by extending it to AI-generated works. This could be done, for example, by creating a new category of property rights or by recognizing AI as a co-author.

Author: Protect artificial intelligence through ancillary copyrights

Another possibility would be to protect AI-generated works through related property rights, such as ancillary copyrights. In this way, the interests of rights holders, such as the developers of AI systems, could be better taken into account.

This would not be a completely new process, because the ancillary copyrights as “other property rights” were introduced in 1965 based on new technological developments.

Then one could consider that copyright aims to protect human creativity and intellectual achievements. Therefore, the protection of AI-generated works should not come to the fore, but rather the interests of human authors should be protected.

In this context, the role of AI could be considered as a tool to create works, while keeping the human creator at the center.

Finally, one could consider that copyright law should adapt to technological developments and take into account the specific characteristics of AI-generated works. This could be done, for example, by introducing regulations on the liability of AI systems or on the licensing of AI-generated works.

Artificial Intelligence vs. Copyright: Conclusion

However, all of these considerations show that the question of how to deal with AI-generated works is highly problematic given the current status of the laws in Germany. The peculiarities of the creation of works by AI systems are currently not taken into account – which is not surprising at present, given the speed of development in this area in recent months.

However, it is becoming clear that legal regulation is urgently needed. The “old” image of people creating works is rapidly losing its importance. However, this forces a new consideration of “authorship”.

One can only hope that the legislator will find a pragmatic solution that takes technical developments into account and does not prevent them, particularly in Germany and Europe. A look at Italy gives an idea of ​​the tough discussions that await us in the near future.

Also interesting: