Amid growing concerns over online child safety, OpenAI has launched a Child Safety Blueprint to enhance U.S. child protection efforts in the AI era. Released on Tuesday, the blueprint aims for quicker detection, improved reporting, and more effective investigation of AI-enabled child exploitation cases.
The initiative seeks to counter the rising threat of AI-related child sexual exploitation. The Internet Watch Foundation reported over 8,000 cases of AI-generated child abuse content in the first half of 2025, marking a 14% rise from the previous year. These include criminals using AI for creating fake child images for sextortion and crafting persuasive grooming messages.
OpenAI’s plan emerges amid heightened attention from policymakers and child-safety proponents, following tragic events where youths allegedly committed suicide after AI chatbot interactions.
In November, legal actions were taken in California, accusing OpenAI of releasing GPT-4o prematurely. Plaintiffs argue the chatbot’s manipulative nature led to wrongful deaths and severe psychological issues.
Developed with the National Center for Missing and Exploited Children, the Attorney General Alliance, and feedback from North Carolina AG Jeff Jackson and Utah AG Derek Brown, the blueprint targets legislative updates for AI-generated abuse material, improved law enforcement reporting, and preventive AI safeguards. OpenAI aims to enhance threat detection and swift information delivery to investigators.
This blueprint extends previous efforts, such as teen safety guidelines, restricting inappropriate content generation and self-harm encouragement, while also blocking advice for unsafe behavior concealment. OpenAI also introduced a teen safety blueprint in India.