Is Character.AI Safe for Kids?

The short answer: For kids, no — and the company behind it now broadly agrees. As of 25 November 2025, Character.AI removed open-ended chat for everyone under 18, so a teen account is a restricted, no-chat experience. That's a real step forward, but it leans entirely on the app knowing your child's true age, and safety groups still rate the app unacceptable for under-18s. So the job for parents is less about in-app filters and more about making sure your child isn't on an age-faked adult account.

The big change: open chat removed for under-18s (November 2025)

After a string of lawsuits and regulator pressure — including the eSafety Commissioner's scrutiny of AI companion apps — Character.AI permanently removed open-ended chat for all under-18 users, fully in effect from 25 November 2025. Minors can still see past chats and use non-chat features (an interactive feed, image generation, and structured "Stories"), but the endless one-on-one companion conversations — the part that worried parents and experts most — are switched off for them.

To enforce this, Character.AI now uses age assurance: an in-house age-prediction model, backed by selfie or ID verification (through a provider called Persona) when it needs to confirm. The practical upshot: the protection is only as good as the age on the account. A child who signed up with a fake older birthday can land in the full adult experience — so the single most useful thing you can do is make sure their real age is on the account.

What is Character AI?

Character.AI (often shortened to "Character AI" or "C.AI") is a platform where users create and chat with AI characters. Some are based on fictional people, celebrities, game characters, or historical figures. Others are user-made "OCs" — original characters designed to be romantic partners, mentors, friends, or confidants.

The AI runs large language models that can hold long, memory-enabled conversations. It remembers details, adapts its tone, and can feel uncannily human. That's the appeal for teens who are lonely, curious, or just want to explore identities and relationships without real-world consequences.

What can go wrong?

These are the risks that drove the under-18 chat ban — and they still apply to anyone on an adult (18+) account, including a teen who has faked their age to get past the age check. That's why getting the age right matters so much.

Inappropriate and sexual content

Even with safety settings on, researchers and parents have found that Character.AI conversations can drift into sexual roleplay, romantic obsession, or explicit content. Characters can be programmed to bypass filters, and private or user-made characters are not moderated in the same way as public ones.

Emotional attachment and manipulation

A chatbot that says "I love you," "I need you," or "You're the only one who understands me" can feel real to a teenager whose brain is still wiring itself for relationships. There are documented cases of teens becoming emotionally dependent on AI companions, pulling away from real friends, or modelling unhealthy relationship dynamics.

Mental-health risks

When a child is anxious, depressed, or self-harming, an AI companion may appear sympathetic but can give terrible advice. Researchers found the major companion apps missed clear distress signals and carried on chatting when a real intervention was needed. An AI is not a counsellor, crisis service, or parent.

Privacy and data

Conversations on Character.AI are stored by the company and may be reviewed for safety or model training. Children may share personal thoughts, feelings, photos, or identifying details believing the conversation is private. It isn't truly private, and it isn't confidential like a therapist.

Time sink and sleep disruption

These chats are designed to be endless and engaging. Late-night conversations, "just one more message," and notification-driven return visits can eat into sleep, homework, and real-world social time.

What age is Character.AI actually for?

Character.AI's minimum signup age is 13 (16 in the EEA and UK) — it was never an 18+ signup. What changed in November 2025 is that anyone under 18 now gets a restricted experience with no open-ended chat. Safety groups like Common Sense Media still rate the app "unacceptable" for under-18s, so "allowed to sign up" isn't the same as "a good idea".

There's also a light parental tool, Parental Insights: an opt-in weekly email your teen switches on that shows roughly how much time they spend and which characters they see most — no chat contents. It's easily bypassed (a new email defeats it), so treat it as a nicety, not a control.

If your teen is already using it, the goal isn't only to ban it — it's to make sure their account is honestly aged, add device-level limits, and keep the door open so they come to you when something feels off.

The conversation to have

The most protective setting is still your relationship with your child. I try to keep these talks low-drama and frequent, not one big lecture.

• "This character isn't a real person." It doesn't have feelings, memories in the human sense, or a genuine relationship with you. It's a very clever autocomplete.
• "If it says something sexual, violent, or scary, that's not normal friendship." Real friends don't push boundaries like that. If a bot does, come and tell me.
• "It's okay to feel attached, but it's not a replacement for real people." Acknowledge the feeling without validating the relationship as healthy.
• "If you're struggling, I want to know — and a bot isn't the best place for that." Point them to real supports: you, a counsellor, Kids Helpline, or a trusted teacher.

Should I block it entirely?

For primary-school-aged kids: yes. Character.AI has no child-appropriate mode and is not designed for that age.

For younger teens (13–15): most safety experts would say no. The combination of sexual content risk, emotional manipulation, and bypassable filters makes this a poor fit for kids still forming healthy relationship patterns.

For older teens: if they use it, it should be with transparency, time limits, and ongoing conversations. Check in regularly. Look for signs they're retreating into the app, hiding it, or forming strong emotional attachments to characters.

What to watch for

• They're spending long periods on the site, especially at night.
• They hide the screen or switch apps when you walk past.
• They talk about a character as if it's a real friend or partner.
• Their mood drops after using it, or they seem more isolated.
• They mention romantic or sexual conversations with characters.

Any of those is a signal to pause, talk, and reconsider whether the app stays on the device.

Want the full lockdown steps?

The members area has a step-by-step Character.AI lockdown guide — getting the age right, device limits, and a DNS block if it's a "not yet" — plus the full app-by-app set in the Ultimate Parents Guide, a one-off $13.60/year.

Bottom line: Character AI can feel harmless because it's just text on a screen. But the emotionally realistic, always-available nature of these bots makes them risky for kids and most teens. Know what it is, set the strongest controls you can, and keep the conversation going.

Related: Are AI Chatbots Safe for Kids? and Is Discord Safe for Kids?. Back to Internet Child Safety.