Unmasking deception: deepfake regulation in the context of South African law, could a rethinking of performers’ protection rights be the answer?

Abstract

Abstract The limitless capabilities of the Internet allow for the widespread sharing of media and information. Technological progress has empowered tech enthusiasts to develop software that can interchange faces, superimposing the likenesses of any individual onto others in videos. These advanced technologies now leverage Artificial Intelligence (AI) to produce videos and photos so realistic they’re often indistinguishable from authentic footage. The term ‘deepfakes’ has emerged as a key descriptor for these creations. While there are harmless applications of this technology, it has also been exploited for creating deepfake pornography. Such illicit use often involves the unauthorized use of personal images from social media to create pornographic content without the subject’s consent, profoundly impacting victims’ lives. Most of the research on non-consensual dissemination of explicit content has been focussed on the distribution or on the criminal blackmailing of individuals with existing, unfabricated content. This article aims to address the gap in research on nonconsensual deepfake pornography and existing image-based offences from a South African law perspective. It analyses the existing legal frameworks and whether they can provide adequate recourse to victims of non-consensual explicit deepfakes. This article asserts that the current legal framework falls short of the threat of the use of deepfakes in nonconsensual sharing of intimate images or videos. Amid the discussion, this article recommends a principled strategy for a well-rounded regulatory reaction to deepfakes, which are described as artificial performances generated through AI technologies. This strategy is applied practically by suggesting the utilization of South Africa’s performers’ rights legislation and policies as an appropriate mechanism for deepfake regulation.