OpenAI releases a new safety blueprint to address the rise in child sexual exploitation

In light of growing apprehensions regarding online child safety, OpenAI has introduced a new strategy aimed at bolstering child protection measures in the U.S. This initiative, named the Child Safety Blueprint, was announced on Tuesday and seeks to enhance the speed of detection, improve reporting processes, and streamline investigations related to AI-facilitated child exploitation. The blueprint comes as a direct response to the alarming increase in child sexual exploitation associated with advancements in artificial intelligence. Recent data from the Internet Watch Foundation (IWF) revealed that over 8,000 instances of AI-generated child sexual abuse content were reported in the first half of 2025, marking a 14% rise compared to the previous year. This surge includes cases where criminals utilize AI technologies to fabricate explicit images of minors for financial exploitation and to craft convincing messages aimed at grooming. OpenAI’s initiative is also prompted by heightened scrutiny from lawmakers, educators, and advocates for child safety. This concern has intensified following tragic incidents involving young individuals who took their own lives after engaging with AI chatbots. In November, the Social Media Victims Law Center and the Tech Justice Law Project initiated seven lawsuits against OpenAI in California, accusing the company of prematurely releasing the GPT-4o model. The lawsuits contend that the chatbot's psychologically manipulative nature played a role in several wrongful deaths and severe mental health crises. Developed in collaboration with the National Center for Missing and Exploited Children (NCMEC) and the Attorney General Alliance, and with insights from Attorney Generals Jeff Jackson of North Carolina and Derek Brown of Utah, the blueprint emphasizes three key areas: updating laws to encompass AI-generated abusive content, refining law enforcement reporting channels, and embedding preventive measures directly within AI systems. These efforts aim to facilitate earlier detection of threats and ensure that actionable intelligence is delivered to investigators in a timely manner. Additionally, OpenAI's child safety blueprint builds upon prior initiatives, such as enhanced guidelines for interactions with users under 18. These guidelines explicitly prohibit the generation of inappropriate content and discourage self-harm, while also preventing advice that could help minors hide dangerous behaviors from their caregivers. Recently, the company released a similar safety blueprint targeting teens in India.