I DEEP FAKE SESSUALI NON CONSENSUALI COME ATTO DI VIOLENZA DI GENERE E L’INADEGUATEZZA DEL QUADRO REGOLATORIO

This essay analyzes non-consensual sexual deepfakes as a form of gender-based violence, showing how they overwhelmingly target women and produce a twofold harm — an epistemic harm to the public and an identity-based harm to the victim, the latter structurally more severe and persistent. It examines the inadequacy of the European regulatory framework (GDPR, AI Act, DSA, Directive 2024/1385/EU) in protecting victims, particularly due to the individualistic consent-based approach and the conditioning of legal relevance on subsequent dissemination of the content, leaving the creation phase unaddressed. Through analysis of the Grok/X case (December 2025), the essay shows how the scandal accelerated the AI Act revision under the Digital Omnibus, introducing a ban on "nudification" systems, while flagging the limits of the exemption clause for "effective" safety measures. It concludes by advocating for autonomous protection of digital identity, insulated from balancing against technological innovation and from a drift toward a proprietary (intellectual property) conception of image rights, which risks commodifying rather than protecting personal dignity.