The Storage Paradox And Algorithmic Disgorgement: Navigating Intellectual Property In Ai Training Data
- IJLLR Journal
- Apr 11
- 1 min read
Isha Singh, LL.M. (Business Law), Amity Law School Lucknow, Amity University, Uttar Pradesh, Lucknow
Dr. Axita Srivastava, Assistant Professor of Law, Amity Law School Lucknow, Amity University, Uttar Pradesh, Lucknow Campus
ABSTRACT
The advent of hyper-realistic generative artificial intelligence and deepfakes has precipitated a profound legal crisis in India, destabilising established conceptions of identity, authorship, and liability. This research paper interrogates the emerging "personality rights gap," demonstrating how current statutory frameworks particularly the Information Technology Act, 2000 and ad hoc judicial injunctions are inadequate against the synthetic misappropriation of an individual's capacity for action. Employing a doctrinal critique and case-study analysis of evolving jurisprudence, including the landmark Anil Kapoor judgment, the paper highlights the erosion of intermediary safe harbour protections and the unresolved intellectual property challenges inherent in AI training data. To address these systemic inadequacies, this paper proposes a novel framework of "Deepfake Torts" under company law. Drawing upon the principle of absolute liability established in Indian environmental law (M.C. Mehta v. Union of India), the research argues for holding corporate AI developers strictly and vicariously liable for harms caused by their inherently dangerous generative models. Furthermore, the paper advocates for the statutory recognition of digital persona rights and the implementation of Algorithmic Disgorgement (model deletion) to ensure the robust protection of human identity in the digital age.
Keywords: Generative AI, Deepfakes, Personality Rights, Deepfake Torts, Algorithmic Disgorgement.
