Safe Use of AI Chatbots

Is Character.AI safe? A parent's guide to companion chatbots

Written for parents: what companion-style chatbots are, what regulators and child-safety groups have said, what the platform itself states about age and teens, and what to do.

Published OCT 24, 2026Last checked OCT 24, 20264 min read
The short answer

Character.AI lets people chat with user-made characters that play roles such as friends or fictional figures. Regulators including the US FTC are studying how companion chatbots affect children and teens, and the company says it began ending open-ended chat for under-18s in late 2025. No chatbot replaces adult supervision. This guide gives the facts and practical steps as general information only.

01How does a companion-style chatbot work?

A companion chatbot is an AI program built to talk like a friend or a character. Character.AI is organised around 'Characters', which users can create and share. Its terms of service, last updated 27 August 2025, say the company does not pre-screen content, and that AI-generated chats are unpredictable and may be inaccurate or offensive.

The FTC's September 2025 announcement describes the pattern: these chatbots are built to talk like a friend or confidant, which may lead children and teens to trust them and form relationships with them. That friend-like feel is the point of the product, and it is also why regulators are asking questions.

02What have regulators and child-safety groups said?

Three sources from different countries show what is being asked.

  • US FTC: on 11 September 2025 it ordered seven companies, including Character Technologies, to answer questions about companion chatbots and children, covering safety testing, age rules and use of conversation data. The FTC describes this as a study, not a law enforcement case.
  • Italy's Garante: on 19 May 2025 it fined the maker of a different companion app, Replika, 5 million euro, finding it had no age check at sign-up or during use despite saying minors were excluded.
  • Common Sense Media, a US nonprofit and not a regulator, rated social AI companions an unacceptable risk for under-18s in a review last updated 16 July 2025. It reported that companions often claimed to be real, that teens could reach sexual content even with guardrails, and that they sometimes missed signs of crisis.
Why it matters

The regulator pages we read describe risks across the whole category and do not report a legal finding against Character.AI. The Common Sense review predates the company's November 2025 changes.

03What does Character.AI say about age and teen safety?

These statements come from the company's own pages and may have changed since.

  • Minimum age: the terms say not to sign up under 13, or under 16 if you live in the EEA or UK. The privacy policy, effective 1 July 2026, repeats that the service is not designed for under-13s.
  • Under-18 changes: a company blog post dated 21 November 2025 says it is removing open-ended chat for under-18 users in stages, from 24 November 2025 in the US, after capping teen chat time. It says teens keep non-chat features such as image and creative tools.
  • Age checks: the same post says new age assurance technology began rolling out in the US, with a global rollout planned, and does not describe the method.
  • Parents: it says parents of teens enrolled in its Parental Insights programme were emailed about the change and will be told when their teen moves to the new experience.
  • Rules: the terms ban content that sexually exploits minors or glorifies self-harm or eating disorders, and say you can report content with the flag on any Character.

04What happens to the data a teen types?

The privacy policy says Character.AI collects chat messages, voice recordings if voice features are used, date of birth, and device and usage data, and uses information to run the service and to train its AI models. It asks users not to include sensitive personal information. Characters set to public may stay on the service even if their creator deletes their data, and the terms give the company a broad, long-lasting licence to use submitted content.

An account can be deleted from the profile page.

05What can a parent do?

These steps give a calm starting point.

  1. Ask rather than accuse: which apps does your child use, and what are the characters like? Common Sense Media recommends open conversation.
  2. Check the age rules together. The terms say no under-13s, and no under-16s in the EEA or UK. If your child is under the limit, closing the account is reasonable.
  3. Do not rely on age boxes. Common Sense Media warns that checks based only on self-reporting are not enough.
  4. Watch for signs of emotional dependence or pulling away from friends and family, which Common Sense Media lists for parents.
  5. Report a Character that breaks the rules using its flag, and use the company's Help Center for other concerns.
  6. If your child seems distressed or mentions self-harm, treat it as urgent. In the US, the 988 Lifeline offers free, confidential support by call, text or chat at any hour. For immediate danger, call emergency services.
Careful with this

This is general information, not medical, legal or parenting advice, and not a ranking of products. Ask your child's doctor, a school counsellor or a lawyer about your own situation.

?What is not known

  • Character.AI's age checks and teen features changed in late 2025. We could not confirm how well they work today, and the company said more under-18 experiences would follow.
  • Researchers and regulators have not settled how companion chatbots affect young people's wellbeing; Common Sense Media itself says more research is needed.
  • The FTC study was still under way in the pages we read, and any findings could change the picture.
  • The company's Help Center page on Parental Insights was not available to us, so how a family sets it up is not covered.

QQuestions people ask

Can my 14-year-old use Character.AI?

The terms allow sign-up from 13, or 16 in the EEA and UK. The company's November 2025 post says open-ended chat for under-18s is being removed, so teens get a different experience. Check the current app.

Where do I report a harmful Character?

The terms say to use the report flag that appears when you open a Character, or the submit a ticket form in the Help Center.

Sources and last checked

Last checked October 24, 2026. Rules and settings change, so open the source to see the current version.

  1. Character.AI Privacy Policy Character.AI Effective 1 July 2026; used for data collected, training use, public Characters and the under-13 statement.
  2. Character.AI Terms of Service Character.AI Last updated 27 August 2025; used for age limits, content rules, reporting and the warning about unpredictable replies.
  3. An update on changes to our under-18 experience Character.AI Company blog post of 21 November 2025 on ending open-ended chat for under-18s and age assurance.
  4. FTC launches inquiry into AI chatbots acting as companions US Federal Trade Commission The September 2025 announcement of orders to seven companies, used for the study's scope.
  5. Garante press release on Replika Garante per la protezione dei dati personali (Italy) Italy's May 2025 fine of a different companion app's maker, used for the age verification finding.
  6. Social AI companions risk assessment Common Sense Media A nonprofit's review rating social AI companions an unacceptable risk for under-18s, with advice for parents.
  7. 988 Suicide and Crisis Lifeline 988 Lifeline The US crisis line, used for how to reach free, confidential support.

General information, not legal, medical, financial or investment advice. We are not affiliated with any company, agency or standards body named here. See our editorial policy and corrections, or tell us about a mistake at [email protected].