ABSTRACT
In both Europe and the United States, courts seek to deny copyright to works developed with Generative AI (GAI) tools, in an effort to maintain the functionality of the status quo. No decision could be more radical – namely, to create a dichotomy of rights between otherwise identical works. GAI tools are being widely integrated into all manner of professional creative technologies – mere buttons to be pushed within a creative workstation as part an author’s process. As such, this dichotomy does nothing less than call into question the copyright status of every artwork from here on in until we can confirm that no GAI tools were used in its production. Yet, disclosure of GAI use tells us absolutely nothing. There is an infinite spectrum of uses of GAI tools, from those that negligibly affect the final work to those that conjure entire works with negligible effort. Being informed that an author has used a GAI tool provides no information whatsoever about that author’s relation to the work itself without granular understanding of their entire creative process. Absurdly, regulatory bodies that wish to enforce this new dichotomy are seeking to undertake an impossible audit of the creative process of every artwork from here-on-in to check if it meets an uncertain authorship threshold.
Hovering a copyright ability question mark over all new works risks destabilising the entire creative economy, as each individual author who wishes to use these popular tools is unsure what rights they hold. In turn, a lack of harmonisation of authorship thresholds worldwide means that creators are unsure where they will hold rights. This dichotomy undermines both of copyright’s pillars, in its inability to protect neither the dignity of creators nor the economic market around them. Rather, each new work may be in the public domain, may be exclusively owned, or may hold different rights in different countries.
Yet even were this dichotomy of rights in otherwise identical works purposively coherent, it is entirely unauditable and unenforceable. In turn, it destabilises and disharmonises the international creative economy blindly without means of asserting its own framework. Without means of enforcement, a copyright framework that denies ownership of GAI-assisted works inherently incentivises denial of use of GAI-tools in order for an author to achieve full rights to their work. As such, redesign of copyright frameworks should not focus on trying to ascertain appropriate authorship thresholds as to when an artwork has had enough human intervention, if it is prima facie clear that it is a literary or artistic work without any confusion as to its stated author. Any copyright framework that seeks to assert a dichotomy of rights in identical works without means of enforcement is a paper tiger – a distraction from the challenges posed to foundational elements of copyright frameworks by new modes of creative production that demand consideration.
Cooper, Zachary, The AI Authorship Distraction: Why Copyright Should Not Be Dichotomised Based on Generative AI Use (July 24, 2024).
Leave a Reply