OpenAI Introduces New Image Classifier and Safety Measures to Combat CSAM
OpenAI has announced a suite of new safety tools designed to detect and prevent the creation or access of child sexual abuse material (CSAM). Central to this initiative is a new image classifier, developed in collaboration with the National Center for Missing and Exploited Children (NCMEC), which can identify abusive imagery—including AI-generated content—with high precision. Early testing indicates the tool significantly reduces reviewer workload by flagging 48% fewer false positives while maintaining detection rates. In addition to technical classifiers, OpenAI is implementing user-facing deterrents within ChatGPT, such as warning screens and temporary account suspensions for users attempting to generate prohibited content. The company is also providing specialized investigative tools and training for law enforcement agencies. These measures come amid increasing regulatory scrutiny and criticism from safety organizations regarding the potential risks AI poses to minors.

Topics: AI safety content moderation CSAM prevention product launch
Read the full article at TechCrunch
Part of the story OpenAI Launches New Classifier to Detect AI-Generated Child Abuse Material (2 articles)