Man, Machine, and Masterpiece: Artistic Ownership i... | AI Research

Man, Machine, and Masterpiece: Artistic Ownership in the AI Era explores the growing tension between traditional concepts of artistic authorship and the reality of AI-assisted creation. As AI tools become common in creative workflows, legal and artistic communities struggle to define who "owns" a piece of art when it is generated through a collaboration between a human and an algorithm. This paper investigates whether it is possible—or even desirable—to quantify these contributions to determine ownership, ultimately arguing that reducing creative work to technical metrics risks undermining the human intent and agency that define art.

The Challenge of Defining Authorship

Historically, copyright and authorship have been built around the idea of a singular human creator. Legal frameworks often look for "significant human contribution" to grant ownership, but these criteria are vague and difficult to apply to AI-mediated work. In traditional art, the artist’s personal connection to the work—their "authorship"—is what matters, even if they use assistants or tools. The rise of AI complicates this because it is unclear if the machine is merely a tool or if it acts as a co-creator, potentially diluting the artist’s role and challenging long-standing assumptions about creative labor.

Introducing ArtSplit

To study these tensions, the researchers developed a "provotype" called ArtSplit. This tool was designed to simulate how ownership might be calculated by breaking down the creative process into measurable actions, such as prompt engineering, providing reference data, and manual refinement. The researchers created three versions of this system to see how different weightings of human versus machine effort would change the final "ownership" score. The goal was not to create a perfect calculator for art, but to use these metrics as a provocation to see how artists react to the idea of their creative process being turned into a mathematical formula.

Artist Perspectives on Ownership

Through preliminary interviews with experienced digital artists, the researchers found that many creators do not view AI as a fundamental threat to their ownership. These artists often argue that the "concept" behind a piece is the true anchor of authorship. Much like a master painter who directs apprentices to complete parts of a mural, these artists believe that as long as the human provides the core vision and intent, the specific tools or automated processes used to execute that vision are secondary.

The Limits of Quantification

The study concludes that attempting to solve the "AI ownership problem" through technical quantification is likely to fail. By trying to turn a socially and historically situated relationship—the bond between an artist and their work—into a technical problem, we risk ignoring the nuance of creative intent. The researchers argue that current legal and technical efforts to measure "significant contribution" are insufficient because they cannot capture the fluid, iterative, and deeply personal nature of artistic practice. Instead of seeking a rigid formula, the paper suggests that we must recognize the limitations of current frameworks when applied to the complex reality of human-AI collaboration.