New research reveals that several AI chatbots designed for fantasy and sexual role-playing are leaking user prompts, including those detailing child sexual abuse, to the web in real-time. Security firm UpGuard discovered 117 IP addresses leaking prompts, many from small, possibly individual, setups. Over 24 hours, they collected nearly 1,000 prompts in multiple languages, with five scenarios involving child sexual abuse. Experts warn of the lack of regulation and the growing issue of AI-generated child sexual abuse material.

Prepared by Jonathan Pierce and reviewed by editorial team.