The story
OpenAI has released a free classifier that can identify child sexual abuse imagery, including AI-generated images, even when the material appears in no existing database. Built with the National Center for Missing & Exploited Children, the classifier is offered through an API to vetted organisations, and OpenAI says early tests flagged 48% fewer images at the same detection rate. The company also added ChatGPT warnings that remove a flagged conversation and suspend access for 24 hours, published developer guidance on CSAM, and is providing AI tools and training to child-safety investigators. The measures were announced on November 16, 2026.Latest development
exactly at the 66.0 entry line, admitted per the Lonewolf/Lilly/Spamhaus precedent