For the first time, a Watch Foundation Internet study (IWF), called to our attention from France Info, identified images of sexual abuse in minors generated by artificial intelligence and transmitted through a chatbot site. Accessible from the “Clara” website, this site is presented as a space for harmless interactions with virtual characters.
But in reality, it allows its users, through a hidden channel, access chatbots that simulate explicit sex scenarios with children, some accompanied by images generated by AI.
Between June and August 2025, the IWF identified 17 reports of this type of content, 94 % of which are “sexual and naked positions.” Mainly they placed children from 11 to 13 years old, even if a particularly worrying case refers to a child of only 7 years. If IWF analysts obviously did not generate images of this type themselves, they have seen that the site and its users actively created this criminal content.
“The progress of AI is rapidly exploited for malicious purposes”
The study emphasizes that these practices are prohibited in the United Kingdom, where it has been carried out and where the authorities can order their suppression. At the same time, a bill is prepared to specifically criminalize the tools of AI designed to produce said content.
The IWF requires a rapid adoption of this text and a stricter regulation of AI, which is incorporated into the safeguards to avoid such derivations. “These chatbots are deliberately created to simulate hateful sex scenarios with children; their existence is unjustifiable,” Kerry Smith also denounced.
She insists on the urgency of the action: “The progress of AI is rapidly exploited for malicious purposes. The new crimes provided for in the bill of crime and the police are necessary, but it is necessary to go faster.”
In the European Union, the debate on children’s pornographic images generated by AI is also in the center of attention, but meets a surprising setback. In addition to that, these contents are part of a legal vacuum, not all member countries wish to supervise and prohibit such content at the European level. Therefore, some countries would prefer to do it at the national level. For the great despair, in particular, of the childhood defense associations that are concerned about the explosion of the availability of these impressive synthetic images. The IWF recalls that these contents could be created thanks to the AI trained by real illegal pornographic content …
Source: BFM TV
