The Algorithmic Authorship Threshold: identifying human contribution in AI-assisted creativity

Abstract Artificial intelligence-assisted creativity increasingly challenges the practical application of EU copyright law by obscuring how human creative contribution should be identified within technologically mediated production processes. While the Court of Justice of the European Union has developed a coherent originality doctrine centred on the author’s ‘own intellectual creation’, existing case law presupposes creative environments in which human agency remains directly observable. In artificial intelligence (AI)-assisted workflows, however, creative autonomy may be fragmented across prompting, iterative refinement, selection and evaluative intervention, leaving courts without a structured method for distinguishing authorship-relevant contribution from mere tool use. This article argues that the principal difficulty is methodological rather than normative. EU copyright law does not require a redefinition of originality; rather, it requires an operational framework capable of applying existing doctrine under conditions of algorithmic mediation. In response, the article develops the Algorithmic Authorship Threshold (AAT), a structured analytical framework that conceptualizes human creative autonomy across three cumulative layers: Creative Intent, Structural Decision-Making and Outcome Shaping. By assessing contribution cumulatively rather than mechanically, AAT seeks to preserve the human-centred foundations of EU copyright law while accommodating contemporary hybrid creative practices. The article further explains how authorship-relevant contribution may be rendered visible in adjudication through a methodological approach focused on evidentiary intelligibility rather than procedural formalism. Through analysis of EU doctrine and selected national traditions, particularly Germany and France, the article demonstrates that AAT neither harmonizes outcomes nor lowers the originality threshold. Instead, it offers a doctrinally faithful and operationally workable method for enhancing consistency, transparency and legal certainty in AI-assisted creativity disputes across the European Union.