Identification Of Digital Security Threats In The Trend Of AI-Based Photo Manipulation: A Systematic Literature Review

Abstract

Advances in AI-based technologies have given rise to a new trend in the creation of manipulated images by combining facial photographs with specific prompts. However, behind what is often perceived as a form of entertainment, this trend presents potential digital security threats to privacy, identity, and personal data that remain largely unrecognized due to limited public literacy and persistent digital divides. Therefore, this study aims to map the digital security threats associated with the growing trend of AI-generated photo manipulation using a Systematic Literature Review (SLR) approach. Employing a qualitative methodology, data were collected through searches of the Scopus database and Publish or Perish, covering publications from 2020 to 2025. Following the identification, screening, and inclusion stages adapted from the PRISMA guidelines, 239 studies were initially identified, of which 18 were selected for further analysis based on their relevance and credibility. The findings reveal that digital identity threats associated with AI-based photo manipulation include privacy violations, non-consensual image manipulation, biometric data theft, identity misuse, and the exploitation of digital footprints. These threats can facilitate various forms of digital fraud, distort representations of past events, and enable sexual harassment targeting vulnerable groups, particularly women and children.