Skip to the main content
PeabodyMake a free story

AI sites and apps

11 "kid-safe" AI chatbots and assistants, and what kid-safe really means for each

"Kid-safe" is not a standard. For one product it means a parent can read every chat; for another it means the bot refuses some topics; for a third it only means the terms say 13+. In 2026 the big lesson was that guardrails leak: Common Sense Media rated Meta AI "unacceptable" for teens, called ChatGPT for teens an unacceptable risk in October, and Character.AI ended open-ended chats for under-18s. This list explains what each option actually does, so you can choose with open eyes.

We judged each by parent visibility, content filters, age terms and independent reviews. Terms checked in October 2026; they changed several times this year. Checked 2026-10-24. Nobody paid to be here and nothing is an affiliate link.

  1. PinwheelGPT

    Pinwheel · Free up to 20 chats/month; about $5.99/month unlimited (Oct 2026) · Ages Kids and tweens with a parent account

    Kid-safe here means filtered answers built for kids plus a parent dashboard showing every conversation. The most supervised option we found. It uses OpenAI models underneath.

  2. Khanmigo

    Khan Academy · About $4/month or $44/year (Oct 2026) · Ages Kids and teens via a parent account

    Kid-safe means focused on schoolwork, parents can read chat history, and it flags worrying messages. It is a tutor, not a friend. Strong safety design.

  3. Kidgeni

    Kidgeni · Free tier; paid plans · Ages 5-12

    Mostly an art studio, but kids can ask questions about the world. Answers draw on curated kid-safe sources, and sensitive topics are flagged to talk about together.

  4. Gemini via Family Link

    Google · Free · Ages Under 13 with Family Link; 13+ on own account

    Kids under 13 can use Gemini only through a parent-managed Family Link account. Google says it does not train on their data, and parents can turn it off.

    Watch for: Common Sense Media rates Gemini for under-13s as high risk. Parents do not see chats.

  5. ChatGPT with parental controls

    OpenAI · Free; paid tiers · Ages 13+; parent consent under 18

    Teen accounts get stricter content rules; age prediction can switch them on. Linked parents can turn off memory, set quiet hours and get distress alerts, but cannot read chats.

    Watch for: No sign-in is needed to ask questions. Common Sense's Oct 2026 report urged limiting it to adults.

  6. Microsoft Copilot

    Microsoft · Free; paid tiers · Ages 13+

    Kid-safe here means a 13+ minimum age (higher in some countries) and extra limits on experimental features. Not designed for children. Parents get no chat view.

  7. Snapchat My AI

    Snap · Free · Ages 13+

    A chatbot pinned inside Snapchat. Through Family Center, parents can stop My AI from replying to their teen, but cannot see private chats. Parents can see time spent.

    Watch for: Lives inside a social app; your teen must accept Family Center.

  8. Meta AI

    Meta · Free · Ages 13+

    Built into Instagram, WhatsApp and Facebook. Meta paused teen access to its AI characters in January 2026 while it built parental controls; the assistant itself stayed on.

    Watch for: Common Sense Media rates Meta AI unacceptable for teens.

  9. Character.AI

    Character.AI · Ages 13+ by terms; no open chat under 18

    Since November 2025, users under 18 cannot have open-ended chats with characters, after lawsuits involving teens. Age checks use behavior signals and ID checks. Teens are pushed to story tools.

    Watch for: Companion-style bots are the riskiest kind for kids. We would avoid it.

  10. Claude

    Anthropic · Free; paid tiers · Ages 18+

    Kid-safe here simply means not for kids: consumer Claude accounts are for adults 18 and older. Fine for parents preparing lessons. Teens should not sign up.

  11. Grok

    xAI · Ages Check current terms

    Listed as a warning. Common Sense Media published a 2026 review calling the Grok chatbot not safe for teens. It is built into X, where many teens already scroll.

    Watch for: Has companion characters and loose filters. Keep kids off it.

More lists

All 112 lists →