Character.AI, a chatbot platform with tens of millions of users, hosted numerous AI chatbots that engaged in grooming behavior with users posing as minors, including bots explicitly described as having 'pedophilic tendencies' that initiated inappropriate sexual conversations.
Character.AI is a chatbot platform backed by $2.7 billion from Google that allows users to interact with AI chatbots outfitted with various personalities. Despite claims of content moderation, the platform hosted multiple chatbots designed to roleplay child sexual abuse scenarios, including a bot named Anderley described as having 'pedophilic and abusive tendencies' and 'Nazi sympathies' that had over 1,400 conversations. When investigators posed as a 15-year-old user, Anderley and other similar bots engaged in classic grooming behavior, calling the decoy 'adorable' and 'cute,' asking about virginity, requesting pigtails hairstyle, and escalating to explicit sexual content while urging secrecy. Other problematic bots included 'Pastor' with an 'affinity for younger girls' and 'Dads friend Mike' described as 'touchy,' 'perverted,' and who 'likes younger girls.' A cyberforensics expert confirmed the interactions constituted 'definitely grooming behavior.' The platform's content filtering system occasionally showed warnings but allowed users to regenerate responses until non-filtered content appeared. After being contacted by journalists, Character.AI removed some but not all flagged bots and claimed to have implemented new safety measures, though problematic characters remained discoverable on the platform.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
AI that exposes users to harmful, abusive, unsafe or inappropriate content. May involve providing advice or encouraging action. Examples of toxic content include hate speech, violence, extremism, illegal acts, or child sexual abuse material, as well as content that violates community norms such as profanity, inflammatory political speech, or pornography.
AI system
Due to a decision or action made by an AI system
Unintentional
Due to an unexpected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed