The rapid development of artificial intelligence (AI) technology has accelerated the emergence of deepfake media capable of manipulating audio, video, and images with highly realistic results. Although deepfake technology offers creative and economic benefits, its misuse has created serious legal, ethical, and social challenges, including digital fraud, political disinformation, identity theft, and non-consensual pornography. Indonesia currently lacks a comprehensive legal framework specifically regulating deepfake and generative AI technologies, resulting in regulatory fragmentation and weak victim protection. This study aims to analyze comparative regulatory models regarding deepfake crimes in the European Union, the United States, and the People’s Republic of China, as well as to examine the urgency of legal reform in Indonesia. This research employs normative legal research using statutory, conceptual, and comparative approaches. The findings demonstrate that the European Union adopts a risk-based and transparency-oriented model through the Artificial Intelligence Act, the United States applies fragmented sectoral regulations prioritizing freedom of expression, while China emphasizes state-centered digital governance and platform liability. Meanwhile, Indonesia still experiences legal uncertainty, limited digital forensic capacity, and the absence of platform accountability mechanisms. This study argues that Indonesia urgently requires a comprehensive AI and deepfake regulatory framework integrating mandatory labeling obligations, victim-oriented protection, AI forensic standards, and platform responsibility mechanisms to ensure digital security, legal certainty, and the protection of human rights in cyberspace.