The FBI has charged a US man with creating more than 10,000 sexually explicit and abusive images of children, which he allegedly generated using a popular artificial intelligence tool. Authorities also accused the man, 42-year-old Steven Anderegg, of sending pornographic AI-made images to a 15-year-old boy over Instagram.
The boom in generative AI has led to the widespread creation of nonconsensual deepfake pornography, which has targeted anyone from A-list celebrities to average, private citizens. AI-generated images and deepfakes of minors have also circulated among schools, in one case leading to the arrest of two middle school boys in Florida who created nude images of their classmates. Several states have passed laws against the non-consensual generation of explicit images, while the Department of Justice has said that generating sexual AI images of children is illegal.
Stable Diffusion, which is an open-source artificial intelligence model, has previously been used to generate sexually abusive images and modified by users to produce explicit material. A report last year from the Stanford Internet Observatory also found that there was child sexual abuse material in its training data. Stability AI, which created Stable Diffusion, has said it forbids the use of its model for creating illegal content.
Stability AI, the UK company behind the wide release of Stable Diffusion, said that it believed the AI model used in this case was an earlier version of the model which was originally created by the startup RunwayML. Stability AI claimed that since it took over the development of Stable Diffusion models in 2022, it has implemented more safeguards in the tool. The Guardian has contacted RunwayML for comment.
