Once described as a utopian invention, artificial intelligence has now become a widespread means of creating child sexual abuse material. Since the release of generative AI tools, this technology, marked by its lack of transparency, has left governments and law enforcement worldwide troubled by how this criminal content is produced and distributed.
As AI bots crawl the internet for content to train their models, openly available photos, videos, and other real media—often shared to celebrate significant events—are being incorporated into vast datasets. Human Rights Watch reported that at least 170 images of Brazilian children, originally posted on photo blogs and private platforms, were included in the LAION dataset, which trains large open-source language models like Stable Diffusion.
This project aims to research and investigate the scope and impact of the epidemic of AI-generated child sexual abuse material. It seeks to explore the reasons behind its creation and the mechanisms involved, rather than merely accepting it as an inevitable consequence of new technologies.
