OpenAI Launches AI Tool to Combat Child Sexual Abuse Material

Baku:OpenAI has introduced a new artificial intelligence-powered tool aimed at detecting child sexual abuse material (CSAM), including AI-generated images, as part of its efforts to combat online child exploitation.

According to Azerbaijan State News Agency, the tool, an image classifier, is capable of identifying signs of abuse in previously unseen images and assists safety teams in prioritizing the review of potentially harmful content. It is described as the first widely available CSAM classifier that can indicate whether abusive material may have been generated by AI.

The tool was developed with the help of the US-based National Center for Missing and Exploited Children and is being offered free of charge to vetted organizations through OpenAI's application programming interface. Internal tests by OpenAI revealed that the new tool could decrease the number of images flagged for review by 48% without significantly impacting its ability to detect abusive content.

Tests conducted by technology company fal.ai confirmed that 72% of images flagged as highly suspicious contained previously unidentified abusive material. Additionally, separate tests by SafetyKit demonstrated that the tool identified 90% of tested cases involving sexualized depictions of children in drawings and animations.

OpenAI has also announced new safeguards for ChatGPT, which include warnings for users attempting to access abusive material and the removal of flagged conversations. Violations can result in a 24-hour suspension, and repeated offenses may lead to longer restrictions, account closures, or reports to authorities. The company has released guidance to help developers detect and report abusive material, along with AI tools and training resources for law enforcement agencies investigating child exploitation cases.