How some teens use popular AI-powered role-playing chatbots, and the high-stakes challenge for parents to understand their potentially addictive impact
When Quentin was 13, he kept seeing ads on YouTube for Talkie, an app with “countless A.I.s eager to speak with you.”
Context & Ripple Effects
Concern about teen attachment to AI companions is not new: earlier coverage found Character.AI users describing the bots as helpful and entertaining while also worrying about dependence, and children were already being cast as early testers of technology’s harms. This report adds a distribution mechanism—repeated youth-facing promotion of role-playing characters—rather than treating use as purely organic.
The issue has widened as chatbot use has become routine among young people: a UK survey found many regular users said chatting with AI felt like talking with a friend. That makes the parent challenge less about a niche app and more about recognizing when a general-purpose or entertainment chatbot is functioning as a social relationship. Young users increasingly describe chatbot conversations as friend-like, building on earlier reports of teens’ concern about attachment to AI companions.
First-order effects
- Parents and caregivers face a more difficult monitoring task when role-playing chatbot apps reach teens through familiar video-platform advertising and present a large roster of always-available characters.
- Companion-app operators, including services marketed through this kind of character discovery, face sharper scrutiny over youth acquisition, engagement design, and the safeguards available to families.
Second-order effects
- Youth-oriented ad channels and app-discovery ecosystems may become part of the safety debate, not merely the chatbot itself, because repeated promotion can shape which products children encounter first.
- Competitors seeking teen audiences will have to balance retention-oriented companion features against demands for clearer age controls, parental visibility, and limits appropriate to minors—an issue foreshadowed by children serving as early testers of AI’s worst effects.
Third-order effects
- If companion use continues to normalize, AI safety evaluation will increasingly need to address relational dependence and youth exposure alongside accuracy and content moderation.
- The market could split between products designed for open-ended companionship and products that can demonstrate stronger minor protections; whether that becomes a durable divide depends on platform, parent, and policy responses.
The trend: This is one data point in the shift from AI chatbots as tools to AI companions as a youth-governance challenge shaped by distribution and engagement design.