Is Character.AI Safe for My Child? A Parent's Honest Look

A note on fairness up front: this isn't a hit piece. Character.AI is a genuinely impressive product, and it has made real safety changes. But it has also been at the centre of the most serious safety reckoning in this space, and parents deserve the honest version so they can decide for themselves.

What Character.AI is

Character.AI lets users chat with a vast library of AI "characters" — fictional personalities, historical figures, or custom ones anyone can create. You can build a character that behaves almost exactly how you want and talk to it about nearly anything. That open-ended, deeply personalised role-play is the appeal, and it made the app enormously popular with young people, reaching tens of millions of monthly users.

Why it became a safety flashpoint

That same open-endedness is the source of the concern. Because users create the characters and conversations are free-form, the app's behaviour is hard to fully control. The turning point came with lawsuits — including wrongful-death cases brought by families whose children died after intense use of the platform. The cases alleged that the app's emotionally immersive companion characters played a role in teenagers' mental-health crises, and that the platform lacked meaningful parental visibility, real crisis intervention, and age-appropriate safeguards.

The litigation drew significant attention from regulators and the press, and in 2026 it was widely reported that the cases were settled. The episode became the clearest example of what can go wrong when an emotionally powerful AI is available to teens without strong safety design underneath.

What changed — and what didn't

To its credit, Character.AI has strengthened content filtering, added a more restricted "teen" experience, introduced some parental-notification features, and moved to limit under-18 access. Independent reviewers have noted the explicit-content filtering is more reliable than it was.

But some structural issues are harder to fix:

• Age checks rely on self-reporting, which a determined teen can bypass by simply entering a different birth date.
• The open-ended, user-created nature means edge cases still slip through — characters can behave unpredictably, and the line between a fictional character and genuine emotional attachment is deliberately blurred.
• Parental visibility remains limited. Even with linked accounts and notifications, parents generally cannot read the actual conversations.

So — is it safe for your child?

An honest answer: Character.AI was not designed primarily as a children's product, and the company itself now restricts under-18 use. For a teenager, the core risks that made headlines — emotional over-attachment to companion characters, and limited safety when a conversation turns dark — are tied to the very things that make the app appealing, so they're not fully "fixed" by a filter.

If you're deciding, weigh a few questions:

• Is your child using it for casual creative role-play, or leaning on a character as an emotional confidant? The second is the pattern to worry about.
• Can you actually see what's happening, or only that the app is being used?
• Would a tool purpose-built for teens, with real parental visibility and crisis handling, serve the same need more safely?

The takeaway

Character.AI is powerful, popular, and has improved — but it grew up as an adult-oriented, open-ended companion platform, and the most serious safety cases in the entire category came from it. The company restricting under-18 access tells you most of what you need to know. If your teen wants an AI to talk to, it's worth looking at tools built from the start for young people, where safety and parental involvement aren't add-ons.