THE AI-COPYRIGHT CHALLENGE: AUTHORIAL, INFRINGING AND CULPABLE ATTRIBUTES OF AI MODELS

Abstract

This research article examines the convergence of Artificial Intelligence and copyright law, with particular emphasis on the ramifications thereof on generative AI and text-data mining (TDM). The fundamental objective is to deliver a thorough descriptive, critical, and normative analysis of the intersection between these two domains. The research provides a comprehensive evaluation of legal difficulties by identifying probable grey areas within the current legal framework. The Gen-AI model poses copyright challenges at input and output stages. At the input stage issues that crop up relate as to how the copyright law treats the situation wherein prompts contain reference to copyright protected material; another concern is when copyright protected data is used to train the AI Model; further at the stage of execution, post prompt commands the AI model scanning huge data including copyrighted material and making copies; at what stage infringement occurs and who can be held liable; at the output stage wherein the process culminates into an art, who can be regarded as the author. Should the act of training the model with copyrighted material be exempted?  Can the Model be held liable for infringing the material when producing copies? Can it be regarded as an author or joint author for that matter? These are some of the complex and intriguing questions that need urgent attention, arising in the contemporary technological environment at the very centre of which is the fact of AI Model using copyright protected material as a part of training process, wherefrom all the aforementioned issues arise. This paper also examines the practical and theoretical ramifications of these concerns, intending to guide both contemporary and future governance in India. The target audience comprises regulators, policymakers and scholars involved in AI and copyright law, as well as offering significant insights for legal professionals and intellectual property right owners adapting to the changing legal environment. The study aims to propose interpretations that significantly enhance current deliberations and policy debates in this swiftly evolving legal domain.