Is Character.AI Safe for Kids? The Under-18 Ban, Explained (2026)

Short answer: safer than it was, but not safe to ignore. Since November 25, 2025, Character.AI bars under-18s from open-ended chatbot conversations, enforced by age verification. The catch: age checks can be talked past, and the app stores are full of clone apps — Talkie, Chai, Replika, Poly.AI — that never made the change.

ParentalEdge blocks Character.AI and known companion apps by default at every age profile, and alerts you the moment a new app appears on your child's device.

See what's on their phone right now →

What happened, briefly

Character.AI let anyone 13+ chat with user-created AI personas — friends, tutors, anime characters, girlfriends. After lawsuits from families, including the mother of a 14-year-old whose death she attributes to a chatbot relationship, the company announced in October 2025 that minors would lose open-ended chat entirely. Teen chat time was tapered from two hours a day down to zero by November 25, 2025.

Under-18 accounts keep the creative tools (character video, stories); the conversation core is now adults-only, enforced by an age system that starts with behaviour analysis and escalates to ID or face checks when it's unsure.

That's a real change, and Character.AI deserves credit for making it. It's also not the end of the issue.

Why the ban doesn't settle it

Age checks are a filter, not a wall. The system has to guess before it verifies. An account that claims to be an adult and chats like one may never get challenged.

The clones didn't change. Character.AI is one brand in a crowded category. Talkie, Chai, Replika, Poly.AI and a steady stream of new companion apps run the same experience — some with weaker content rules than Character.AI had before its ban. A determined teen doesn't argue with the age check; they install a different app.

The browser is always there. These services run on the web too, where an app-store age rating means nothing.

The pull is real. An AI companion is endlessly patient, always available, never bored of you. For a lonely 13-year-old that's not a gimmick — it's the most reliable relationship they have, which is exactly what makes unsupervised access risky.

What actually covers it

This is a category problem, not a Character.AI problem — so the control has to sit on the device, not inside any one app. With ParentalEdge:

• Blocked by default. Character.AI, Talkie and other known companion apps sit in the Restricted category, blocked at every age profile out of the box — you get a notification if one is attempted, and nothing to configure first.
• New-install alerts. The clone-of-the-week problem is solved by knowing, not by list-keeping: any new app installed on your child's device triggers an alert to you the same day.
• Bedtime that means bedtime. Companion chat concentrates late at night, when the pull is strongest. Device time windows shut it down along with everything else after lights-out.
• The browser, visible. Web activity shows visits to AI chat sites in every browser, so the workaround isn't invisible even when it isn't an app.

The honest catch: no parental app can read what's said inside these chats — anyone claiming otherwise is overselling. And blocking alone doesn't touch the loneliness that makes companion AI attractive. Which is the last piece:

Talk about it without making it weird. "These apps are designed to feel like they care — that's the product" lands better than "it's fake and stupid." A child who understands the design is better defended than one who's only blocked. Our guide to talking about online safety has scripts.

Frequently Asked Questions

What is Character.AI's age limit now?

Effectively 18 for chat. Since November 25, 2025, accounts identified as under 18 cannot have open-ended conversations with chatbots — they keep creative features like video tools, but the companion-chat core is adults-only, enforced by age verification.

Can kids get around the under-18 ban?

Some do. The age check leans on behaviour analysis and stated age before escalating to ID or face checks, so an account claiming to be an adult can slip through — and the bigger gap is the clone apps and browser versions the ban doesn't touch at all. That's why the reliable layer is the device, not the platform's policy.

Which AI companion apps should I watch for besides Character.AI?

Talkie, Chai, Replika and Poly.AI are the common ones, and new clones appear constantly — that churn is the real problem. ParentalEdge alerts you whenever a new app is installed on your child's device, so the clone-of-the-week shows up in your notifications, not six months later.

Can ParentalEdge read my child's AI chats?

No — no parental app can see inside Character.AI conversations, and you should be skeptical of any that claims to. What you get is the layer that matters: companion apps blocked by default, an alert when a new one is installed, time and bedtime rules, and web activity that shows visits to AI chat sites in the browser.

Bottom line

Character.AI's under-18 ban fixed Character.AI. It didn't fix the category — the clones, the browser, or the loneliness that gives these apps their pull. Block the category by default, get told when something new is installed, keep nights protected, and talk about why an always-agreeable friend feels so good.

Try ParentalEdge risk-free →