OpenAI announces a new image classifier designed to detect potential child sexual abuse material (CSAM), including images generated by AI. Multiple outlets report that the system is meant to help human reviewers and child-safety investigators identify harmful imagery more quickly and reduce the volume of content they must examine.

Alongside the classifier, OpenAI adds additional safety measures: ChatGPT warnings and developer guidance, plus support intended for child safety investigators. Several reports describe the classifier as being developed with assistance from the National Center for Missing & Exploited Children, and say OpenAI is working with partners such as the Tech Coalition and ROOST to reach more organizations.

Outlets differ mainly in emphasis and included details. Some foreground the “stricter interventions” and how ChatGPT responds to misuse, while others focus on the classifier’s role in improving workflows for investigators and reviewers. One outlet also highlights specific performance and enforcement figures, while others do not provide those numbers.