Criminal Liability for Deepfake Cybercrime: Evidentiary Challenges and Regulatory Gaps in Indonesian Law

Abstract

The rapid advancement of artificial intelligence has enabled the emergence of deepfakes synthetic audio-visual content generated using machine-learning models such as Generative Adversarial Networks. While deepfake technology offers beneficial applications, its misuse in cybercrimes such as identity fraud, extortion, and others raises serious concerns in Indonesia due to the absence of explicit legal regulation. This study aims to analyze the criminal liability associated with deepfake-based cybercrime by examining evidentiary challenges and identifying regulatory gaps within the Indonesian Criminal Code and the Electronic Information and Transactions Law. Employing a qualitative library research methodology, this study synthesizes primary legal sources, academic literature, and authoritative institutional reports. Findings reveal that deepfake cybercrime presents substantial evidentiary difficulties, particularly in authenticating electronic evidence and maintaining chain-of-custody integrity, as deepfake content often lacks detectable traces of manipulation. Furthermore, the Criminal Code and Electronic Information and Transactions Law provide only general provisions for forgery, defamation, pornography, and electronic manipulation, without addressing AI-based content fabrication or synthetic impersonation. This regulatory vacuum restricts law-enforcement capability, creates inconsistent legal interpretations, and weakens victim protection. The study concludes that Indonesia requires explicit statutory definitions, specialized criminal provisions for deepfake offenses, and standardized forensic guidelines to ensure legal certainty.