THE LEGAL IMPLICATIONS OF DEEPFAKE TECHNOLOGYCOPYRIGHT INFRINGEMENT AND CRIMINAL LIABILITY

Deepfake technology, grounded in generative adversarial networks (GANs) and latent diffusion models, has emerged as a powerful instrument of misinformation, identity theft, non- consensual intimate imagery (NCII), and commercial fraud. This paper conducts a doctrinal, comparative, and empirical examination of deepfake-related legal issues, focusing on two principal domains: copyright infringement and criminal liability. Drawing upon statutory frameworks, case law, and emerging legislation across the United States, United Kingdom, European Union, Australia, India, and select Asian jurisdictions, the paper argues that existing legal regimes are insufficiently equipped to address the harms generated by synthetic media.
The paper examines the technological foundations of deepfake generation, the copyright dimensions of AI training data, authorship of AI-generated outputs, performer rights, platform liability, and criminal offences ranging from NCII to fraud, electoral interference, and AI- generated child sexual abuse material (CSAM). A comparative jurisdictional analysis reveals fragmented and inconsistent regulatory responses. The paper concludes by proposing a multi- tiered regulatory framework combining technical provenance standards, copyright reform through extended collective licensing, bespoke criminal offences, reformed platform liability rules, and international treaty coordination.
Keywords: Deepfake, Synthetic Media, Generative Adversarial Networks, Copyright Infringement, Criminal Liability, Artificial Intelligence, Regulatory Framework, Digital Identity.