<![CDATA[A new type of cybercrime known as “deepfakes,” which has the ability to manipulate video and audio content in a nearly lifelike manner, has emerged as a result of advances in artificial intelligence (AI) technology. On the other hand, this raises controversy with conventional criminal law doctrine, which stipulates that there must be an element of criminal intent (mens rea) on the part of the individual. The purpose of this study is to identify changes in the concept of mens rea resulting from the autonomization of AI, examine the issue of criminal liability for deepfakes under current Indonesian positive law, and propose an ideal framework for reconstructing the principle of fault for the future. The research method used in this article is normative legal analysis or a literature review, employing a statutory approach to examine relevant legislation, a conceptual approach, and a general case-based approach, with the analysis of legal materials conducted through qualitative normative reasoning using deductive reasoning. The research results show that the autonomous features and black-box nature of artificial intelligence create a disconnect between human subjective intent and the machine’s automated outputs. This leads to a conceptual crisis and a multi-actor liability dilemma among developers, users, and distributors of manipulative content. Furthermore, Indonesia’s current cyber regulations remain highly reactive and lack formal standards for attributing culpability in AI-generated crimes. Therefore, this article concludes by proposing a reconstruction of the principle of culpability through the application of a risk-based approach that shifts the paradigm from proving individual, anthropocentric intent toward an objective assessment of systemic risk. Thus, to ensure certain and fair digital law enforcement, progressive national criminal law reform and the standardization of digital forensics are necessary.]]>
