deepfake

In the era of rapid advances in artificial intelligence and computer graphics, it is difficult for an untrained observer to be able to avoid, as well as recognise as such, an altered image, and now increasingly also an altered video. Known colloquially as "deepfakes," a portmanteau of "deep learning" artificial intelligences and "faked imagery", these seemingly seamlessly altered videos present challenges for notions of authentic representation, and much has already been written about their potential applications to and influencing of political discourses. Still, several aspects of "deepfakes," potentially made manifest both through high-end editing software and open access mobile applications, remain critically underexamined.
Most evident among these is perhaps the instantiation of "deepfake pornography," which relies on the digital superimposition of real people’s, almost exclusively women’s, faces and voices onto pornographic videos, predominantly for the consumption of male users. Often, the superimposed images and sounds come, or are alleged to come, from images that are considered part of the public domain, have been posted publicly by the individuals depicted or ones in possession of copyright, and in other ways allow for transformative use. The implication is, perhaps, that women’s bodies are to be seen as physical objects that, in an era of the incresing accessibility of image- and video-altering software, may as well be digitally recreated so as to be consumed in a way that circumvents any pesky discussions of consent. It is therefore necessary to take a closer look at deepfake technologies along with the exploitative, often violently mysoginistic as well as cisheteropatriarchal and white supremacist social and legal practices that commonly underlie their predominant uses.