This study delves into the intricate relationship between artificial intelligence (AI) generated works and copyright law, focusing on the European Union, the USA, and Japan as major jurisdictions. It explores whether AI-created works are eligible for copyright protection and examines the concepts of authorship and originality within the context of AI-generated content. Despite efforts to harmonise copyright laws, differences persist among these jurisdictions. The EU leans towards protecting works within the Berne Convention’s scope, while the US requires human authorship for copyright eligibility, as established in legal precedents. Similarly, Japan emphasises creativity as a prerequisite for copyright protection, setting a higher threshold compared to other jurisdictions. However, a central challenge arises in assessing human input in AI-generated works, particularly with opaque systems like neural networks. The traceability of human contribution and its creative distance from AI-generated output pose significant hurdles in determining copyright ownership and protection in the evolving landscape of AI technology.