Criminal Regulatory Approaches to Deepfake-Related Offenses: Focusing on the Crime of Fraud

Abstract

As a representative form of AI-generated synthetic media, deepfake technology-characterized by its high degree of realism and increasing accessibility-has expanded the scope of legitimate technological applications while simultaneously providing novel tools and pathways for fraud-related crimes. Centered on the offense of fraud, this article systematically examines the role and functional mechanisms of deepfake technology within the criminal chain, identifying the associated criminal risks across key stages such as the illicit acquisition of personal information, content fabrication, and the execution of fraudulent schemes. The study finds that deepfake-enabled fraud exhibits distinctive features, including low technical barriers to entry, a modularized criminal chain, increasingly precise and multidimensional methods of deception, and the diffusion of harmful effects into the broader system of social trust. These characteristics pose significant challenges to existing criminal law frameworks, particularly with regard to the identification of criminal subjects, the classification of forms of accomplice liability, the allocation of platform responsibilities, and the legal characterization of relevant conduct. On this basis, the article proposes a systematic regulatory framework from three dimensions-legislative refinement, judicial response, and comprehensive governance. This includes promoting the interpretation and adaptation of the elements of fraud crimes, establishing an intelligent trial assistance system, implementing a dual-constraint mechanism for deepfake content labeling and informed consent, and the exploration of a collaborative “platform–government” regulatory model. These proposals aim to provide theoretical support and institutional reference points for the prevention and punishment of deepfake-enabled fraud.