This roundtable will open discussions around the ethical, political, regulatory and technological dimensions of consent in an environment of synthetic and generative media. We suggest that consent is emerging as a defining theme in the altered media environment enabled by deep-learning techniques. Deepfake videos have been used in satirising the powerful, in reimagining histories, in conscripting women into non-consensual pornography, and in resurrecting the dead (Ajder & Glick 2021, Chesney & Citron 2019, Gregory 2022, Paris & Donovan 2019). Synthetic media are not just about new ways of making meanings, but about challenges to settled understandings of how both meanings and media get made (Winger-Bearskin et al 2022), and these processes are tied to new forms of contested data practices (McCosker 2022).

In every case the first line of critique is about the lack of consent from those depicted (Meikle 2023). Generative media tools such as ChatGPT and DALL-E raise questions of intellectual property, and the consent of those who created the model-training content or who may be represented within it. From the harvesting of personal photos as AI training data, to art and style imitation, to the involuntary participants in deepfake porn — each of these issues centres consent. Yet consent is a contested concept. While predominantly linked to notions of justice and agency, its limits are increasingly interrogated by feminist, queer and critical scholars who point to moments when consent becomes an alibi or is assumed by default (e.g. the tech industry small print ‘effective consent’ model, cf Tiidenberg 2018).

Initial participants will offer introductory ideas and provocations on the intersection of generative media and consent from their own research before opening the discussion up to establish connections between diverse digital developments that elide, demand or challenge issues of consent, and to consider possible solutions.