A modern tale of Frankenstein? How to regulate non-consensual sexually explicit AI-generated deepfakes in the Metaverse

Abstract

This thesis examines the emerging issue of non-consensual sexually explicit deepfakes (sexfakes) in the context of the Metaverse. The research focuses on two key issues: sexfakes as a gender-based phenomenon, and the applicability of current regulations on sexfakes in the Metaverse. The thesis takes a euro-centric approach and therefore analyses the current regulatory frameworks established by the Council of Europe (CoE) and the European Union (EU) on sexfakes and the Metaverse. The study reveals that while progress has been made in addressing sexfakes, current protections remain incomplete, leaving women inadequately safeguarded. The Metaverse presents further unique challenges in the regulation and enforcement of sexfakes, raising complex questions about identity, legal personality, avatar ownership, harm, perpetrator accountability, jurisdiction, and the right to be forgotten. This thesis argues that the EU's and CoE's stance that no new instruments are necessary may underestimate the transformative potential of the Metaverse, especially concerning gender-based cyber violence. It proposes to enhance current frameworks to specifically address sexfakes in the Metaverse by taking a proactive regulatory stance so that the Metaverse could be transformed into a safe and inclusive digital space for all. Keywords: artificial intelligence, gender-based violence, deepfakes, non-consensual sexually explicit deepfakes, Metaverse, European Union, Council of Europe.