SexAI
A scientific article on generative AI in adult content · 2025
SexAI is a scientific article written in the first year of Media Engineering at HEIG-VD, together with Ana Viola, Ivan Kubikov and Léana Guex. The exercise: pick a digital-society topic and treat it in a genuine scientific article format (structure, tone, and some twenty cited and verified sources).
We picked an uncomfortable but very real subject: what generative AI is doing to adult content. The article works on three levels. On the technical side, GANs and text-to-image models make hyperrealistic content production accessible to anyone, with no technical skills. The ethics come next, with non-consensual deepfakes that overwhelmingly target women, training datasets contaminated with illegal material, and a creeping dehumanization of sexuality. The economics are just as stark: behind OnlyFans' 6.6 billion dollars sits a median creator income of roughly 150 dollars a month, and fully AI-generated avatars are starting to push real creators out.
We wanted to show that AI's most urgent questions, consent, dignity, cross-border regulation, are at their sharpest exactly where people prefer not to look. The full article can be downloaded below.