Does it matter if agentic AI is an author? effort and Nigerian Copyright Law

Abstract

Agentic AI complicates copyright by unsettling the basis on which authorship is attributed, particularly where effort bears on originality and protection. This article asks what makes effort legally authorial, examining Nigerian copyright law, whose originality threshold requires effort to give a work original character, alongside comparative doctrine. That formulation reflects a common‐law tradition attentive to labour, skill and judgment, distinct from the European emphasis on the author's own intellectual creation and free and creative choices. The article argues that neither the quantity of labour involved in production nor the operational agency attributed to an AI system establishes authorship by itself. What matters is the relation between the claimed effort and the determination of copyright‐relevant expression. This difficulty is especially apparent where systems generate music with limited human intervention. Agentic production thus exposes a broader problem of legal attribution. Copyright may continue to seek an author even where production no longer coheres around one. The article describes the unsettled condition surrounding this shift as nescience, arguing that it counsels legal restraint rather than inaction. Where no qualifying human authorial contribution to the resulting expression can be established, copyright protection should be withheld, leaving fully AI‐generated expression in the public domain.