Copyright vs. Artificial Intelligence: Examining the legal Challenges of AI Training and Authorship Under the Indian Law

The rapid diffusion of generative artificial intelligence has unsettled two foundational assumptions of Indian copyright law that a work must trace back to a human author, and that the material used to build a creative system was lawfully obtained. This article examines the statutory architecture of the Copyright act,1957 , particularly the definition of author under Section 2(d) and its 1994 extension to computer generated works against the backdrop of judicial reasoning on originality developed in cases such as eastern book company v D.B. Modak and R.G. Anand v Delux Flims.It then turns to the more urgent and commercially consequential question of whether training large language models on copyrighted text amounts to infringement ,using the Delhi High Court’s 2026 order in ANI Media Pvt Ltd v Open AI Inc as a lens. The article argues that Indian law is currently ill-equipped on both fronts to deal with machine generated creativity and machine consumed data, and its proposes a set of legislative interventions drawing on comparative practice in the United Kingdom, the United States and Singapore.