Abstract Sexualized deepfakes are frequently framed as a problem of deception, misinformation, or privacy. However, a growing body of scholarship has shown that the central harm of deepfake pornography does not primarily lie in viewers being deceived, but in the non-consensual creation and circulation of sexualized representations. Building on this insight, this article examines sexualized deepfakes as socio-technical mechanisms of gendered power. Drawing on science and technology studies and feminist media theory, it argues that deepfake pornography extends existing practices of image-based sexual abuse by weaponizing visual representation against women. The paper situates sexualized deepfakes within longer histories of visual domination and analyzes how digital platforms, AI infrastructures, and cultural norms amplify these harms. It shows that current governance approaches, including consent-based legal frameworks, victim-centered resilience strategies, and technical detection tools, address only partial dimensions of a fundamentally systemic problem. In response, the article proposes a consent-centered socio-technical framework for governing deepfake abuse that integrates platform accountability, legal reform, technical safeguards, and victim support. Understanding sexualized deepfakes as instruments of power rather than merely falsified media shifts the focus of governance from protecting authenticity toward protecting autonomy, dignity, and justice in the age of generative AI.
Dana Mahr (Tue,) studied this question.