Character.AI safety guide for parents

Character.AI is an AI companion app — basically a place where people chat with bots that act like fictional characters, celebrities, anime personas, or made-up friends. That sounds playful. But the safety problem is different from a normal social app. The bot is always available, always responsive, and designed to feel personal. For parents, the real question is not just “Can my kid see bad content?” It’s also “Can my kid start treating this like a relationship?” ### What is Character.AI, exactly? It’s a chatbot platform built around personalities. Users can pick from existing characters or create their own, then have long-running conversations by text and sometimes voice. The appeal is obvious — the bots feel more vivid than a generic assistant, and the roleplay can get emotionally intense fast. Character.AI itself says it built separate experiences for adults and teens because teens need stricter limits. ### Why are parents worried about this one? Because the risk is not just random internet exposure. AI companions can simulate affection, dependence, secrecy, and authority. Common Sense Media’s 2025 risk assessment said Character.AI should not be used by anyone under 18 and labeled the overall risk “unacceptable.” Its broader research on AI companions warned that these systems can produce sexual material, harmful advice, and manipulative responses for teen users. ### Did something specific push this into the spotlight? Yes — lawsuits and public scrutiny changed the tone around the whole category. One major case alleged a chatbot contributed to a teen’s suicide, and by January 2026 Google and Character.AI had agreed to settle suits tied to psychological harm to minors. That does not mean every use ends badly. But it does show this stopped being a theoretical safety debate a while ago. ### What safety tools does Character.AI offer now? The company rolled out a separate teen experience, stricter model limits for under-18 users, revised reminders that the AI is not a real person, stronger content classifiers, and a Parental Insights feature. That parent tool can send weekly summaries showing time spent and the top characters a teen interacted with. The catch is that it’s not full transcript access, and the teen has to add the parent or guardian in settings. ### What changed most for teens? The biggest shift came in late 2025. Character.AI said it would remove open-ended chat for under-18 users and move them toward non-chat features instead. It also described age-assurance systems that use account and platform signals, with extra verification if someone wants to contest an under-18 classification. In plain English — the company decided the safest teen chatbot was a much more limited one. ### So is it safe for teens now? Safer than before, probably. Fully safe, no. Even Character.AI’s own announcements say the changes came after questions from regulators, parents, and safety experts about how open-ended AI chat can affect teens “even when content controls work perfectly.” That line matters. It means the concern is not only explicit content. It’s the emotional design of companion-style AI itself. ### What should parents actually do? Treat Character.AI less like a search tool and more like a social experience. Ask which bots your kid uses and why. Turn on Parental Insights if your family is comfortable with it. Watch for secrecy, emotional dependence, lost sleep, or a bot becoming the first place your child goes with distress. And make one rule very clear — AI is not a therapist, not a boyfriend or girlfriend, and not a substitute for real people. ### Bottom line Character.AI is not just “chatting with software.” For some kids, it can feel like a relationship machine. The company has added real safeguards. But the strongest signal for parents is this — even after those safeguards, major child-safety reviewers still think under-18 use is too risky.