Regulating AI-generated Image-based Sexual Abuse: How Platform Guardianship and Legal Deterrence Shape Offending Proclivity

• In Event: Promoting Safety in Online and Offline Spaces: The Importance of Activity and Structural Features

Abstract

As AI-generated image-based sexual abuse (AI-IBSA) becomes increasingly prevalent, understanding the structural conditions that enable or deter its perpetration is critical. This study employs an experimental vignette design to assess how key situational factors – platform guardianship (regulatory presence), certainty of punishment (perceived risk of detection and legal consequences), and severity of punishment (legal categorization of AI-IBSA) – shape individuals’ proclivity to engage in AI-IBSA. By manipulating these factors, this study experimentally evaluates whether variations in platform oversight, legal enforcement and detection, and perceived punishment influence individuals’ willingness to engage in AI-IBSA perpetration. Findings from this research will contribute to discussions on the effectiveness of regulatory interventions and inform policy efforts aimed at preventing AI-IBSA.

Authors

• Sohee Jung, University of South Carolina

• Hunter Boehme, University of South Carolina