aceEmerging evidence
More
Children are making AI sexual images of classmates, a review of research argues
Study title: Explaining peer-generated AI-CSAM: A criminological analysis of youth offending, gendered harm and technology-facilitated sexual abuse
- Evidence type
- Commentary, opinion, or narrative review
- Sources
- 1 sourcePeer-reviewed
- Published
- 2 min read
- Truvace record
- TRV-2026-2516cite or check it
Study title: Explaining peer-generated AI-CSAM: A criminological analysis of youth offending, gendered harm and technology-facilitated sexual abuse
A review published in Criminology & Criminal Justice examines how children and young people are using AI to create sexual images of their peers, including modified images and deepfakes that fall under the category of child sexual abuse material. It treats this as a form of technology-facilitated gender-based violence, with women and girls the main targets.
Read the study abstract
Artificial intelligence has transformed the landscape of sexual harm, particularly towards women and girls, by enabling the rapid creation of synthetic hypersexualised material, including peer-generated modification images and deepfakes, falling under the category of child sexual abuse material (AI-CSAM). While research has focused primarily on adult offenders, far less attention has been given to the growing involvement of children and young people in producing AI-generated sexual imagery of peers. This review…
The problem
A literature review argues that AI tools now let children and young people quickly make synthetic sexual images and deepfakes of peers, mostly targeting women and girls, and that adult-focused research has left this youth offending understudied.
In brief
The authors say most research has concentrated on adult offenders, leaving the growing involvement of young people in producing such imagery far less studied. They draw on moral disengagement theory, social learning theory, information gap theory and feminist analyses of image-based abuse to explain the behaviour.
The review places the problem within a safeguarding system that spans the family, the school and the national curriculum, and argues that weakness at any one level weakens protection overall. It proposes an Ecological Safeguarding Model for coordinated prevention across those domains.
Main points
- The review defines peer-generated AI-CSAM as synthetic hypersexualised material, including modification images and deepfakes, made by young people about their peers.
- It frames the problem as technology-facilitated gender-based violence, with women and girls particularly affected.
- It says research has mostly examined adult offenders, with far less attention on children and young people who produce such imagery.
- Theories used include moral disengagement, social learning, information gap and feminist analyses of image-based abuse.
- The proposed Ecological Safeguarding Model links family, school and national curriculum responses.
- The authors identify gaps in current safeguarding practice and implications for policy, education and future research.
The full story
The article is a narrative review that synthesises behavioural, criminological, feminist and sociotechnical literature rather than reporting new measurements. Its aim is a theoretical framework for explaining why young people create AI sexual imagery of peers.
It combines several theories: moral disengagement and social learning to explain how offending behaviour develops and is justified, information gap theory, and feminist work on image-based abuse that treats the harm as gendered.
On responses, the authors argue that safeguarding works across a multi-level ecology of family, school and national curriculum, and that a failure at one level undermines the rest. They say coordinated intervention across these domains can strengthen prevention, and they call for trauma-informed, gender-aware and developmentally appropriate approaches.
Sources
1 on file- Criminology & Criminal Justicedoi:10.1177/17488958261467387
What readers say