OpenAI launches classifier to help child abuse investigators identify AI-generated imagery
OpenAI has announced a classifier intended to help child sexual abuse investigators distinguish actionable harmful material from the large volume of reports they receive.
The tool can detect abusive imagery involving children and indicate whether it was generated by AI, according to OpenAI. The company says the classifier is designed to help investigators prioritize material that can be acted on quickly, particularly as synthetic content adds to the burden on reporting systems.
OpenAI is also offering free access to vetted organizations, publishing guidance to help developers detect this type of content, and supporting training for law enforcement on how AI can assist investigations. The announcement comes amid broader calls for stronger restrictions and safety controls on AI tools.