Adverse human rights impacts of dissemination of nonconsensual sexual deepfakes in the framework of European Convention on Human Rights: A victim-centered perspective

Generative artificial intelligence systems have advanced significantly over the past decade and can now generate synthetic but highly realistic audio, photo, and video, commonly referred to as deepfake. Image-based sexual abuse was the first widespread (mis)use of deepfake technology and continues to be the most common form of its misuse. However, further (empirical) research is needed to examine this phenomenon's adverse human rights implications. This paper analyses the potential adverse human rights impacts of the dissemination of nonconsensual sexual deepfakes in the framework of the European Convention on Human Rights and argues that the dissemination of such deepfakes can hinder the rights protected by the Convention. These include the right to respect for private and family life, as nonconsensual sexual deepfakes can undermine data protection, harm one's image and reputation, and compromise psychological integrity and personal autonomy. Additionally, such deepfakes can threaten freedom of expression by creating a silencing effect on public watchdogs, politicians, and private individuals. Finally, nonconsensual sexual deepfakes can impair the economic and moral rights of pornography performers by abusing their work and bodies to abuse others without authorization and compensation. These findings highlight that the Council of Europe member states must fulfil their obligations to provide effective protection against this technology-facilitated, gender-based, and sexual violence.