Man, Machine, and Masterpiece: Artistic Ownership in the AI Era

This paper investigates the tension between traditional notions of artistic authorship and the integration of AI-driven systems in creative work. The authors argue that current legal and regulatory frameworks—particularly those in the EU—struggle to account for distributed agency in AI-mediated art. To explore this, the researchers developed 'ArtSplit,' a provotype designed to explicitly quantify human and AI contributions across various stages of the creative process, such as prompting, data input, and refinement. By presenting these quantitative breakdowns to artists, the study aimed to provoke critical reflection on whether such metrics can or should define artistic ownership.

Through preliminary interviews and the use of the ArtSplit provotype, the authors found that professional artists often reject the idea that ownership can be reduced to a set of measurable actions. Participants emphasized that ownership is primarily rooted in the 'underlying concept' of a work, rather than the specific labor steps involved in its execution. The study reveals that artists view their relationship with AI as similar to historical practices of delegating tasks to apprentices or assistants. Consequently, attempts to quantify 'significant human contribution' are seen as misaligned with the reality of creative practice, where the artist's identity and intent remain the primary markers of authorship, regardless of the tools used.

The impulse to solve the 'AI authorship problem' through technical quantification risks diluting the social and historical foundations of art. By treating ownership as a technical problem to be solved by algorithms, regulators may inadvertently undermine the artist's role as the conceptual anchor of their work. This research highlights the necessity of moving beyond simple input-output metrics when discussing the future of creative labor and legal recognition in the age of generative AI.