Konstruksi Hukum Pidana Dalam Penanggulangan Kejahatan Siber Berbasis Teknologi Deepfake di Indonesia

Abstract

The development of Deepfake technology presents new challenges for Indonesian criminal law, particularly in the field of cybercrime. Deepfake, as an artificial intelligence creation, can realistically manipulate audio and video, making it susceptible to misuse for fraud, non-consensual pornography, defamation, and digital identity theft. This study focuses on the construction of Indonesian criminal law in addressing Deepfake-based cybercrime. The research employs a normative juridical approach combined with a comparative study. The data consist of primary legal materials (the ITE Law, the Criminal Code, the Personal Data Protection Law, and court decisions), secondary legal materials (scholarly journals, books, research reports), and tertiary legal materials (legal dictionaries, encyclopedias). Data were collected through library research and documentation, and analyzed using descriptive-comparative techniques. Data validity was ensured through source triangulation and cross-verification. The findings show that the construction of Indonesian criminal law in handling Deepfake-related crimes still relies on extensive interpretation of the ITE Law (Articles 27, 28, 35) and relevant provisions of the Criminal Code on morality, fraud, and defamation. However, this construction remains inadequate, as there is no explicit provision regulating AI-based content manipulation, leading to legal uncertainty in terms of offense elements, evidentiary standards, and criminal liability. Nevertheless, there is potential for adaptation through analogical interpretation, the broad jurisdiction principle of the ITE Law, and general criminal sanctions. To strengthen this legal construction, it is necessary to establish specific regulations defining Deepfake in legal terms, provide graded sanctions based on impact, ensure victim protection (including the right to be forgotten), and impose preventive obligations on digital platforms.