On Friday, Europol announced that 25 people – including three Swedes – had been arrested in a major crackdown on AI-generated material depicting sexual abuse of children.
Thomas Andersson, an investigator at the children's rights organization Ecpat Sweden, notes that AI-generated abuse material is a major problem that has only grown in recent years. This is evident at Ecpat Sweden, which has a web-based hotline where the public can report all forms of sexual exploitation of children.
99 percent of all tips we receive concern images or videos on the internet. And in our work on this hotline, we have seen a steady increase, roughly from 2021-2022 and onwards, of AI-generated material. I would say an increase both in terms of the number of images and the quality, i.e. the photorealism of the images.
"Large quantities every day"
Thomas Andersson says he often gets asked if AI-generated images are not a "better" alternative.
When we talk about AI-generated abuse material, it's not one thing, but several different things. Very often when we see AI-generated material, we actually see what is called AI-manipulated material. That is, taking pictures of existing victims, children who have been subjected to abuse that has been photographed, and manipulating those images into new positions or new actions, he says, and continues:
And the consequences for the child are the same as other abuse material – that is, when it spreads on the internet, it's as if the abuse never ends.
Used for blackmail
Thomas Andersson says that perpetrators also use AI tools to create manipulated images of real children. Images that are then used to blackmail the children.
They pressure them for money, otherwise the images will be sent to people they know.
Thomas Andersson further notes that the consumption of abuse material – whether it's real photos and videos or AI-generated – drives demand for more material. This means that more children risk being subjected to abuse.
