Ó Ciardha, C., Buckley, J., & Portnoff, R. S. (2026). AI-generated child sexual abuse material: What’s the harm? AI & SOCIETY.

Link: https://doi.org/10.1007/s00146-026-02932-y

Open Access: Yes

Notes: AI-generated child sexual abuse material (AI-CSAM) is one of the key issues that communities and policymakers have sought to address. Nonetheless, a key debate remains over how to compare AI-CSAM with traditional CSAM, with some arguing that the synthetic nature of the former lessens its impact compared with the latter. In this article, Ó Ciardha and colleagues argue that the dangers of AI-CSAM should not be downplayed. Here, the authors advance seven arguments: 1) in many cases, AI-CSAM depicts real children; 2) offenders explore AI content to coerce, groom and extort; 3) it normalizes and desensitizes CSAM; 4) it is a gateway to offending; 5) it generates new venues for peer exploration without consent; 6) it complicates targets’ protection; and 7) it incentivizes production and monetization. To summarize the dangers, they argue that: “The customizable nature of AI CSAM means that the content can be rendered more violent, sadistic, and otherwise extreme than any underpinning CSAM used in its creation. This customizability also lends itself to the creation of material that is more emotionally engaging for users, for example through the use of chatbots and other multimodal AI tools. This capacity reinforces concerns that synthetic content contributes to escalating patterns of consumption.”