Criminal Aspects of Identifying Deepfakes

The contemporary criminal landscape in the Russian Federation is undergoing a paradigm shift driven by evolving operational methods. It demonstrates a significant surge in offenses that utilize artificial intelligence and AI-generated synthetic media (deepfakes). Conducting effective investigations into these crimes requires a precise comprehension of the markers and detection methodologies necessary to establish robust evidentiary verification in court. Based on a comprehensive review of scientific literature and expert reports, these forensic markers can be categorized by source into image, audio, and video data. Their identification involves an intersection of specialized forensic disciplines, including computer, phonographic, phototechnical, portrait, and video-technical examinations. Detecting signs of manipulation relies on identifying the dynamic features of a subject’s altered physical appearance alongside idiolect anomalies within communications, which can be verified through forensic identification examinations. An integrated approach and a multidisciplinary, complex forensic examination are required given the multi-layered complexity of synthetic AI-generated falsifications.