Authorship in artificial intelligence-generated works: Exploring originality in text prompts and artificial intelligence outputs through philosophical foundations of copyright and collage protection

Abstract

The advent of artificial intelligence (AI) and its generative capabilities have propelled innovation across various industries, yet they have also sparked intricate legal debates, particularly in the realm of copyright law. Generative AI systems, capable of producing original content based on user-provided input or prompts, have introduced novel challenges regarding ownership and authorship of AI-generated works. One crucial aspect of this discussion revolves around text prompts, which serve as instructions for AI systems to generate specific content types, be it text, images, or music. Despite the transformative potential of AI-generated works, the legal landscape remains fragmented, with disparate jurisdictional interpretations and a lack of uniform approaches. This disparity has led to legal uncertainty and ambiguity, necessitating a nuanced exploration of originality, creativity, and legal principles in the context of text prompts and resulting outputs. This article seeks to contribute to the ongoing debate by delving into the complexities surrounding AI-generated works, focusing specifically on the originality of text prompts and their correlation with resulting outputs. While previous literature has extensively examined copyright issues related to AI, the originality of text prompts remains largely unexplored, representing a significant gap in the existing discourse. By analysing the originality of text prompts, this article aims to uncover new insights into the creative process underlying AI-generated works and its implications for copyright law. Drawing parallels from traditional creative works, such as collages, the article will assess how legal principles apply to AI-generated content, considering philosophical foundations as well as copyright principles, such as the idea-expression dichotomy. Furthermore, the article will explore the divergent approaches taken by different jurisdictions, including the United Kingdom, United States, and European Union, in determining originality in the context of copyright law. While refraining from providing definitive answers, the article aims to stimulate critical thinking and dialogue among stakeholders. By offering fresh perspectives and insights, it seeks to enrich the discourse surrounding the copyrightability of AI-generated works and pave the way for informed policy decisions and legal interpretations. The article aims to contribute valuable perspectives to the ongoing debate on copyright and AI, shaping the future trajectory of intellectual property law in the era of artificial intelligence.