Encrypted, Synthetic, and Unaccountable: Generative-AI Sexual Imagery on Messaging Platforms, the Limits of Traceability, and the Attrition of Justice

Generative AI has made it quick and cheap to create sexually explicit images of real people, and these images spread easily through encrypted and semi-private apps such as WhatsApp and Telegram. This paper is a structured narrative review of recent evidence from hotline reports, police data obtained through freedom-of-information requests, regulatory documents, and technical research. It examines how large and fast-growing the problem is, why encrypted platforms make detection and tracing harder, how few cases end in a legal outcome, and which responses have the strongest support. The evidence suggests that women and girls are the main targets, that victim dropout and limited police capacity matter as much as encryption, and that the best-supported responses work without weakening encryption. The paper ends with recommendations, including standard reporting of case outcomes.