THE ROLE OF HUMAN CONTRIBUTION IN AI-GENERATED WORKS

Abstract

Although generative artificial intelligence has made it simpler to produce wellpolished texts and images, it has also reduced the visibility of the human author's role. The user might offer an idea, a prompt, source material, detailed instructions, repeated feedback, or make substantial edits. However, these actions do not have the same legal significance. The article looks at how copyright law can tell the difference between human intellectual creation and the functioning of a generative system. It begins by examining the EU's originality criterion based on the Court of Justice's judgments in Infopaq, Painer, Football Dataco, Cofemel, Levola, and Brompton Bicycle, and the 2025 Mio and Konektra cases. It then compares the relevant laws of Germany, France, Spain, Italy, and Ireland and looks at the first Czech ruling concerning an AI generated image. The analysis also takes into account Albanian legal scholarship and Law No. 35/2016, even though Albania is not an EU member state, because these are relevant to the broader European discussion. The main argument of the article is that authorship should be evaluated in relation to the specific expression in question. A court should identify the person's independent creative decisions, check whether those decisions are present in the final work, and grant protection only for the human contribution that results. The cases demonstrate why a simple rule based on counting prompts would be insufficient: a prompt can remain an unprotected idea, while human activities such as composition, selection, revision, or arrangement can result in protectable expression. This method allows for creative work produced with AI to be recognised while still attaching copyright to demonstrable authorship. The article ends by proposing a practical framework for demonstrating contribution and making clear the legal implications of mixed human and machine production.