ARTIFICIAL INTELLIGENCE AND COPYRIGHT: WHO OWNS CREATIVITY IN THE AGE OF GENERATIVE AI?
Generative artificial intelligence is changing how creative works are made.
Texts, images, or music… , AI tools can now create different types of content in a matter of seconds.
These technologies offer surely unlimited opportunities for innovation, however, from another side, they raise complex copyright questions:
Can AI generated work be protected by copyright? If so, who owns the IP rights in these creations? The AI tool? the AI company? the user? or nobody at all?
Copyright law has traditionally been based on the idea that protected work must come from human intellectual effort and originality. This principle of authorship appears in both national and international copyright laws, which assume that an author is a person whose creativity leads to protected work.
For decades this idea worked without much controversy since creative works were always linked to human authors. While advances in technology have changed the tools used in creativity, they haven’t fundamentally challenged the belief that a human creator stands behind every protected work.
However, the rise of GENERATIVE AI tools is putting these traditional legal rules to the test. AI now can create music, novels, visuals and other types of content without or with limited human intervention.
As a result, copyright questions arise. For example, if a marketing agency uses SUNO to generate music, CHATGPT to write advertising copy and to create visuals for a campaign, the campaign may achieve commercial success, but who owns the copyright for these creations?
It’s clear that the level of human involvement in the creative work is crucial in determining if it qualifies for protection. If an AI generated work has significant human input that led to final result, it may still be protected. Conversely, if the work is generated entirely by AI with little or meaningless human input, its copyright statutes become much more uncertain.
It’s important to note that the current legal landscape faces many issues concerning AI generated content. For instance, in 2025, the U.S copyright office stated that “copyright does not extend to purely AI generated material or material where there is insufficient human control over expressive elements” [1] emphasizing that AI generated work must involve human intervention to be protected.
This protection is illustrated by the Thaler v. Perlmutter case, concerning an artwork “a recent entrance to paradise” generated autonomously by an AI system known as the “creative machine”. Thaler acknowledged that he had not created the artwork himself but argued that as the owner of the AI system, he should nevertheless own the copyright in the resulting work.
The U.S copyright office refused registration, finding that a work created entirely by a machine without human intervention cannot be protected. This position was upheld by the U.S district court in 2023[2] and by the court of appeals [3]
However, this approach is not universally followed. The U.K provides an interesting alternative. Under section 9(3) of the copyright, designs and patents act 1988, in the case of a literary, dramatic, musical or artistic work that is computer generated, the author is considered to be “ the person by whom the arrangements necessary for the creations of the work are undertaken” [4] the act further defines a computer-generated work as one generated by computer in circumstances where there is no human intervention
This provision is particularly relevant to the current debate because it shows that copyright law can, at least in principle, attribute authorship even where no traditional human author directly creates the work
Nevertheless, its application to modern generative AI remains open to discussion in practice, identifying the person who made the “necessary arrangements” may not always be straightforward
Who then should be regarded as having made the necessary arrangements for the AI creations: the developer who designed the AI system? or the user who provided the prompts?
While the UK has rules, even if unclear, for computer generated work, many jurisdictions, including Tunisia, still lack clear rules on AI generated work. Existing copyright law were created before generative AI emerged and often do not clarify if purely AI created content can be protected, or who, if anyone, owns the rights.
Given the current legal uncertainty around AI generated works, we may need to think about revising existing copyright laws.
Some potential paths could include:
One way would be to keep the traditional requirements of human authorship but clarify how much human contribution is needed for an AI assisted work to gain copyright protection. This would ensure AI remains a tool, with protection depending on how much creative control the user had over the final result.
However, distinguishing between a simple prompt and a genuine creative contribution can be tricky.
Another possibility is to follow the U.K model and create specific rules for works generated without human contribution. Rights could go to whoever made the necessary arrangements.
A third option would be to establish a separate right for purely AI work, giving them a unique form of legal protection without equating them to traditional human creations. Yet, lawmakers would still need to decide who benefits from this right and how long the protections last.
Lastly, purely AI generated work could remain outside copyright protection while continuing to protect works involving enough human creativity. This would uphold the traditional foundations of copyright law but could cause concerns for businesses that invest in creating and using AI generated content.
The choices made today will shape not only the future of copyright, but also the evolving relationship between human creativity and AI.
As AI continues to transform the creative process, copyright law will need to adapt in a way that embraces innovation while preserving the essential role and value of human creators.
Written by: Noura Feki “Legal advisor”
One world IP – Tunisia office
