Regulating Synthetic Realities: A Comparative Analysis of Global Legal Responses to AI-Generated Deepfakes

Abstract

The proliferation of AI-generated deepfakes has prompted diverse regulatory responses across jurisdictions, revealing fundamental tensions between technological innovation, individual rights, and societal protection. This article provides a comparative analysis of legal and policy frameworks addressing deepfake-related harms in the United States, European Union, China, South Korea, and Pakistan, examining how different legal traditions and institutional contexts shape regulatory approaches. It identifies three emerging regulatory paradigms: application-based regulation targeting specific high-harm uses; life cycle-based regulation addressing synthetic media at different phases of creation and dissemination; and subject-based regulation focusing on platform and service provider responsibilities. The analysis reveals that while no jurisdiction has fully resolved the challenges posed by deepfakes, proactive legal frameworks, specialized institutional capacity, and preventive strategies significantly enhance response effectiveness. Conversely, reliance on technology-neutral cybercrime provisions, fragmented institutional approaches, and limited forensic capacity substantially constrains regulatory efficacy, particularly in developing jurisdictions. The article concludes by identifying best practices and transferable lessons for jurisdictions seeking to strengthen their legal and institutional responses to deepfake-related harms.