Gender and generative AI – Future of women’s safety and deepfake

Abstract

Purpose The purpose of this study examines that over the course of just a few years, deepfakes have become tools of synthetic misogyny, amplifying the scope, pace and anonymity of technology facilitated gender-based violence (TFGBV). Design/methodology/approach This a conceptual, interpretive and future-oriented study that uses causal layered analysis (CLA) as the main analytical method to deconstruct the deepfake-enabled TFGBV at four levels: litany, systemic causes, worldview/discourse and myth/metaphor. The analysis is drawn from feminist futures theory, intersectionality and technofeminism, and relies upon existing studies. Based on the CLA findings, four scenarios for 2025–2040 are designed: two scenarios related to the uncertainty of regulatory effectiveness and two to the uncertainty of the strength of the feminist movement. Findings The findings show that without feminist-centered governance, deepfake ecologies will continue reinforcing gendered inequalities, undermining democratic participation and normalizing digital sexual violence. Finally, the paper proposes a roadmap for feminist artificial intelligence (AI) governance, structural accountability and culture-changing interventions. Research limitations/implications Scenario forecasting (2025–2040) outlines four alternative futures, demonstrating how regulatory strength and feminist mobilization determine the direction of deepfake harm. Practical implications The paper proposes a roadmap for feminist AI governance, structural accountability, and culture-changing interventions. Social implications Findings show that without feminist-centered governance, deepfake ecologies will continue reinforcing gendered inequalities, undermining democratic participation and normalizing digital sexual violence. Originality/value It applies feminist futures studies and CLA approaches to analyze TFGBV as a multi-layered socio-technical issue, not an exception to the technological order.