AI and authorship redefined: towards a global copyright framework for commerce and human originality

Neubauer, Anja M

You are here : University of Gloucestershire > Research > Research Repository

AI and authorship redefined: towards a global copyright framework for commerce and human originality

Tools

Neubauer, Anja M (2025) AI and authorship redefined: towards a global copyright framework for commerce and human originality. PhD thesis, University of Gloucestershire.

Full text not available from this repository.

Official URL: https://www.researchgate.net/profile/Anja-Neubauer...

Abstract

Purpose This thesis explores the way in which generative AI changes the fundamental premises of copyright law and threatens human creativity. It examines legal gaps concerning authorship, training data use, moral rights, and human originality in the context of AI systems that can mimic human expression. The aim is to create a global copyright framework that will ensure innovation and legal certainty, commercial application, and cultural and personal value of human-created works. Design/Methodology/Approach The study is based on a qualitative, interpretivist, and inductive research design. It is a combination of comparative legal analysis in three jurisdictions (UK, US, Germany), a scoping review of academic literature and empirical data from semi-structured interviews with legal experts, AI developers, industry stakeholders, and content creators. The findings are arranged using the MATH–COPE framework, which examines the interaction of four legal aspects: moral rights, authorship, training data, and human originality with four influencing factors: commercialization (the economic impact of AI on creative markets), organisational practice (how AI is used in everyday work), policy and governance (changes in laws and court decisions), and ethical technology (the responsibility of developers and following rules). Findings The research indicates that the current copyright laws are ill-prepared to deal with the semantic and stylistic replication abilities of AI. There is no legal consensus on authorship, there is no clear framework for licencing training data, and there are not enough safeguards for the moral identity of creators. Interviewees were very concerned about attribution failures, lack of transparency in AI development, and the absence of human creators’ compensation mechanisms. Notably, there are no explicit legal protections against the posthumous use or replication of a deceased person’s likeness or style through AI, and only fragmented personality rights to fill the void. The final framework suggests specific reforms: a redefined legal definition of ‘work’, licensing requirements for training data, improved moral and posthumous rights, and enforceable transparency and remuneration. Originality/Value This thesis is one of the first to combine doctrinal analysis, commercial implications, and stakeholder perspectives into a holistic legislative proposal. It presents a modular global framework of eleven articles and operationalizes the MATH–COPE matrix as a replicable tool for assessing law, ethics, and innovation concurrently. The distinction between AI-assisted and AI-generated content, along with the proposed semantic originality threshold, contributes new insights to the international copyright discourse. Limitations The research examines three countries (UK, US, Germany) and the Berne Convention. The EU AI Act, TRIPS and DMCA were not considered to avoid confusion in the discussion. The EU AI Act was not discussed, since its focus on risks does not deal with copyright matters related to training data and human creativity. The Berne Convention gives a better and more effective way to study the global issues related to AI-generated content. Furthermore, the fast development of generative AI could mean that some parts of the proposed law will need to be updated in the future. Overall contribution (to theory/practice) The study shows that copyright laws are not adequately prepared for the rise of AI-generated content. It provides a solid foundation for future international changes, for example by WIPO, to adapt copyright laws and protect human creativity in the age of artificial intelligence. The study offers a structured and legally based answer to the urgent challenges of generative AI. It aids legal scholarship by reinterpreting traditional copyright categories to reflect technological changes and by providing actionable policy instruments for lawmakers, companies, and creators. And it sets the stage for a forward-looking copyright system that preserves human originality without killing innovation.

Item Type:

Thesis (PhD)

Thesis Advisors: