Your child probably meets AI already: chatbots for homework, “companion” apps built to feel like a friend, and AI tucked inside apps they use every day. Each company sets its own minimum age, but a calm, repeated conversation does more than any setting.
Parental controls change settings rather than show you the chats. This page covers what kids meet, the age rules in the companies’ own words, scams and fake images, and how to talk about it.
- What kids meet: chatbots, companion apps, and AI features inside social and messaging apps.
- Age rules differ: several services say 13; Anthropic’s consumer Claude app says 18. Read the rule for the app your child actually uses.
- Controls are a fence, not a camera: they limit what an app does, but mostly do not show you the conversations.
- Scams and fakes: a voice can be copied and a photo can be faked. A family code word and a callback beat trying to spot the fake.
- Talk, and keep talking: it can be wrong, it is not a friend, keep personal details out, and come to me if something feels off.
What kids actually meet
The plain point: AI reaches children in three ways. A chatbot they open on purpose, a companion app built for a relationship, and AI built into apps they already use.
The comparison: think of a chatbot as a library help desk, a companion app as a pen pal who never sleeps, and built-in AI as a new staff member who appeared in a shop you visit every day. Each needs a slightly different talk.
- Chatbots (tools like ChatGPT, Gemini or Claude that answer in plain words). Used for homework, questions and curiosity.
- Companion apps. Built to be a character or friend you chat with over time. These need the most care; see the section below.
- AI inside other apps. For example, Snapchat describes My AI as “an experimental, friendly, chatbot” and warns its answers can be wrong.1 Snapchat also says messages with My AI are kept unless deleted and may be used to improve its products.1
How AI got into your tools shows how these features arrive, often switched on. The eight kinds of AI you already use helps you spot them.
Make a list with your child of every app they use, and mark the ones with AI inside. That list is your starting point.
Age rules, in the companies’ own words
The plain point: each service sets its own minimum age in its terms. It is a rule in a contract, not a guarantee that every user is that age.
The comparison: an age rule is like the “12A” or “PG-13” label on a film. It tells you what the maker intends. It does not stop a child from watching.
| Service | Minimum age | Extra conditions |
|---|---|---|
| ChatGPTOpenAI | Age: 13, or the minimum age in your country. | Also: under 18 needs a parent’s or guardian’s permission. |
| GeminiGoogle | Age: 13 with a personal or school account, or the age in your country. | Also: younger children only through a parent-supervised account, not offered everywhere. |
| Meta AIMeta | Age: 13, or higher where a country requires it. | Also: separate terms in some regions. |
| Character.AI | Age: 13; 16 in the EEA and UK. | Also: the company said it would remove open-ended chat for under-18s. |
| ClaudeAnthropic | Age: 18, or higher where local law requires. | Also: the company says it disables accounts it believes belong to under-18s. |
The sources, in order. OpenAI’s terms: “If you are under 18 you must have your parent or legal guardian’s permission to use the Services.”2 Google’s help page sets 13 for Gemini.3 Supervised access for younger children is not offered in the EEA, Switzerland or the UK.4 Meta’s AI terms set 13.5 Character.AI’s terms set 13, or 16 in the EEA and UK.6 In November 2025 it said it would begin removing open-ended chat for under-18 users in the US first.7 Anthropic’s terms set 18.8 Its help centre describes disabling accounts.9
Terms change. AI chatbot age limits and parental controls keeps the fuller picture, including US and California law.
Look up the rule for each app your child uses, on the company’s own page. Treat it as the company’s floor, not a verdict on your child.
What parental controls do, and do not do
The plain point: controls can switch features off, set hours and send some alerts. Most do not let you read the chats.
The comparison: controls are a fence round the garden, not a camera in the house. They shape where a child can go, not what was said.
- ChatGPT. OpenAI says linking a parent’s and a teen’s accounts does not show the parent the chats.10
- Snapchat My AI. In Snapchat’s Family Center, a parent can stop My AI replying to their teen.11
- Gemini. Google says its extra filters for children are not perfect.4
Set controls up with your child, so they know what is on and why. Then check the setting that decides whether their chats train the AI. What happens to what you type explains it.
Companion apps need extra care
The plain point: an assistant answers questions. A companion app is designed to feel like a relationship, and that is where the main worries are.
The comparison: a homework chatbot is like a calculator. A companion is like a friend who is always free, always agrees and never has a bad day. That can feel lovely, and it can make real friends seem like hard work.
The US Federal Trade Commission opened an inquiry because chatbots can feel like a confidant to young users.12 Character.AI itself said in October 2025 that it would end open-ended chat for under-18s.13
The research is mixed: short-term comfort, with signs of withdrawal over longer periods. AI companions sets out what is known and what is not. A useful question for your child: is it a place to rest before going back to people, or has it become the place?
Ask what they talk to it about, not just how long. If a child is telling an app how they really feel, that is a reason for a gentle conversation, not a ban.
Scams, fake voices and fake images
The plain point: AI makes it cheap to copy a voice or fake a photo. Families need rules that work even when the fake is convincing.
The comparison: do not try to be a better lie detector. Change the locks instead: a code word and a callback work whatever the voice sounds like.
Fake voices. Microsoft researchers showed a voice can be copied from about three seconds of recording.14 Agree a family code word, said out loud and never posted. If a “family member” calls in a panic asking for money, hang up and call back on the saved number. The household guide is a printable version.
Fake images of children. The FBI has warned that ordinary photos are being altered into explicit images.15 These are often used for sextortion: threatening to share an image unless the victim pays or sends more. The same alert advises parents to talk about the risks.15
- Tell your child they will not be in trouble for coming to you, even if they sent something.
- Do not pay and do not send more. Keep the messages as evidence and report it.
- Get it removed. In the US, the National Center for Missing & Exploited Children runs Take It Down for nude or sexual images taken of someone under 18.16 US law also requires platforms to remove reported intimate images within 48 hours.17 Outside the US, use your national reporting service.
Deepfake self-defence, how AI fraud works and spotting AI-generated images, video and voices go further.
Set the family code word this week. And say out loud, now: “If anyone ever threatens you with a picture, come to me. You will not be in trouble.”
How to talk about it, with prompts that help
The plain point: four short talks, repeated, beat one big lecture. AI can even help you prepare them.
The comparison: it is like teaching road safety. You do not explain it once. You say “look both ways” at every crossing until it sticks.
- It can be wrong, confidently. Google tells parents Gemini “can hallucinate and present inaccurate information as factual”.18
- It is not a friend. The same page says Gemini “should not be treated as a friend or used to discuss personal problems or secrets”.18
- Keep personal details out: Google lists “address, school name, or family information”.4
- Come to me if a chat feels wrong. A calm reaction keeps the door open.
Use these prompts yourself. Change the parts in [brackets].
Treat the answers as drafts. Check any setting or rule on the company’s own page, as the AI can be out of date. How to explain AI to someone else has more ideas.
Where to go next
Pick the line that sounds like you.
- “I want every app’s rules and the law.” Read AI chatbot age limits and parental controls, app by app.
- “I want to protect grandparents too.” Print the household guide.
- “My child uses AI for homework.” Share AI for students and read AI in schools.
- “I want to talk to the school.” See what AI for teachers says schools are told.
- “I am new to AI myself.” Start with AI explained for complete beginners.
- “I want the safety checklist.” Open how to use AI safely, or when not to use AI.
Questions people ask
Can parents see their child’s ChatGPT conversations?
No. OpenAI says linking a parent’s and a teen’s accounts does not show the parent the chats.10 Most parental controls switch features off, set hours and send some alerts, rather than letting you read the conversations.
How old do you have to be to use ChatGPT?
OpenAI’s terms set 13, or the minimum age in your country, and anyone under 18 needs a parent’s or guardian’s permission.2 It is a rule in a contract, not a guarantee that every user is that age.
How do I protect my family from AI voice scams?
Agree a family code word, said out loud and never posted. If a “family member” calls in a panic asking for money, hang up and call back on the number you already have saved. Do not rely on judging the voice: a voice can be copied from a short recording.14
What should I do if my child faces a sextortion threat?
Tell them they will not be in trouble for coming to you. Do not pay and do not send more; keep the messages as evidence and report it. In the US, Take It Down from NCMEC can help remove nude or sexual images taken of someone under 18.16
QUICK CHECK · 3 QUESTIONS
Pick an answer to see if you got it. Nothing is recorded.
1. What do most parental controls on AI apps not do?
- Switch some features off
- Let you read your child’s chats
- Set hours when the app can be used
Show the answer
Answer: Let you read your child’s chats. The page calls controls a fence round the garden, not a camera in the house: they change settings but mostly do not show the conversations.
2. A family member calls in a panic asking for money. What does the page say to do?
- Judge whether the voice sounds right
- Send a small amount to be safe
- Ask for the family code word, then hang up and call back on the saved number
Show the answer
Answer: Ask for the family code word, then hang up and call back on the saved number. The page says a voice can be copied, so a code word and a callback work whatever the voice sounds like.
3. What is an app’s minimum age rule, according to the page?
- The company’s rule in its terms, which differs from app to app
- A government check that every user has passed
- The same age for every AI app
Show the answer
Answer: The company’s rule in its terms, which differs from app to app. The page says each service sets its own minimum age in its terms, and it is not a guarantee that every user is that age.
Related on this site: AI Chatbots and Mental Health · AI for Older Adults · How to Learn a Language With AI · How to Make a Meal Plan With AI
- 1Snapchat Support, Staying Safe with My AI, read at source 1 Oct 2026“it’s possible My AI’s responses may include biased, incorrect, harmful, or misleading content.” ↑a“all messages with My AI will be retained unless you delete them and may be used to improve Snap’s products and to personalize your experience”. ↑b
- 2OpenAI, Terms of Use, effective 1 Jan 2026, version for users outside the EEA, Switzerland and the UK, read at source 18 Sep 2026 as recorded on AI chatbot age limits and parental controls: “You must be at least 13 years old or the minimum age required in your country to consent to use the Services.”↑a↑b
- 3Google, What you need to sign in to Gemini Apps, read at source 1 Oct 2026“You must be 13 (or the applicable age in your country) or over to use Gemini with a personal or school Google Account”. ↑
- 4Google, Guide your child’s Gemini Apps experience, read at source 1 Oct 2026“Gemini Apps are not available for supervised accounts in the European Economic Area, Switzerland, and the United Kingdom.” ↑a“they’re not perfect”. ↑b“not to share sensitive personal information like their address, school name, or family information”. ↑c
- 5Meta, Meta AIs Terms of Service, effective 13 May 2026, read at source 18 Sep 2026 as recorded on AI chatbot age limits and parental controls: “you cannot access AIs if you are under the age of 13 (or such greater age required in your country or territory)”.↑
- 6Character.AI, Terms of Service, last updated 27 Aug 2025, read at source 18 Sep 2026 as recorded on AI chatbot age limits and parental controls: “If you are under 13 years old OR if you are under 16 years old and a citizen or resident in the European Economic Area (EEA) or the United Kingdom (UK), do not sign up for the Services”.↑
- 7Character.AI, An Update On Changes to Our Under-18 Experience, 21 Nov 2025, read at source 1 Oct 2026“Starting on Monday, November 24, we will begin removing open-ended chat for under-18 users in the US, with other markets following in short order.” ↑
- 8Anthropic, Consumer Terms of Service, effective 8 Oct 2025, read at source 1 Oct 2026“You must be at least 18 years old or the minimum age required to consent to use the Services in your location, whichever is higher.” ↑
- 9Anthropic, Age assurance on Claude, read at source 18 Sep 2026 as recorded on AI chatbot age limits and parental controls: “we’ll disable accounts based on indicators of minor activity.”↑
- 10OpenAI Help Center, Managing parental controls in ChatGPT, read at source 18 Sep 2026 as recorded on AI chatbot age limits and parental controls: “Linking accounts does not give a parent or guardian access to a teen’s conversations, chat history, or real-time activity.”↑a↑b
- 11Snapchat Support, How do I limit my teen’s access to My AI?, read at source 1 Oct 2026“As a parent enrolled in Family Center, you can choose whether My AI replies to your teen.” ↑
- 12Federal Trade Commission, FTC Launches Inquiry into AI Chatbots Acting as Companions, 11 Sep 2025, read at source 18 Sep 2026 as recorded on AI chatbot age limits and parental controls: chatbots “generally are designed to communicate like a friend or confidant, which may prompt some users, especially children and teens, to trust and form relationships with chatbots.”↑
- 13Character.AI, Taking Bold Steps to Keep Teen Users Safe on Character.AI, 29 Oct 2025, read at source 1 Oct 2026“we will be removing the ability for users under 18 to engage in open-ended chat with AI on our platform.” ↑
- 14Microsoft Research, VALL-E, Jan 2023, read at source 10 Sep 2026 as recorded on Deepfake self-defence: “only a 3-second enrolled recording of an unseen speaker”.↑a↑b
- 15FBI Internet Crime Complaint Center, Malicious Actors Manipulating Photos and Videos to Create Explicit Content and Sextortion Schemes, I-060523-PSA, 5 Jun 2023, read at source 1 Oct 2026“The FBI continues to receive reports from victims, including minor children and non-consenting adults, whose photos or videos were altered into explicit content.” ↑a“Monitor children’s online activity and discuss risks associated with sharing personal content”. ↑b
- 16NCMEC, Take It Down, read at source 1 Oct 2026“Take It Down is a free service that can help you remove or stop the online sharing of nude, partially nude, or sexually explicit images or videos taken of you when you were under 18 years old.” The page does not say whether AI-altered images are covered. ↑a“Take It Down is a free service that can help you remove or stop the online sharing of nude, partially nude, or sexually explicit images or videos taken of you when you were under 18 years old.” The page does not say whether AI-altered images are covered. ↑b
- 17Federal Trade Commission, FTC Begins Enforcing the TAKE IT DOWN Act, 19 May 2026, read at source 22 Sep 2026 as recorded on Deepfake self-defence: platforms must “remove those intimate images, and known identical copies, within 48 hours of a valid request.”↑
- 18Google, Guide your child’s Gemini Apps experience, read at source 1 Oct 2026.↑a↑b
Read at source 1 Oct 2026: Snapchat Support, Staying Safe with My AI and How do I limit my teen’s access to My AI? · Google, What you need to sign in to Gemini Apps and Guide your child’s Gemini Apps experience · Anthropic, Consumer Terms of Service · Character.AI, Taking Bold Steps to Keep Teen Users Safe and An Update On Changes to Our Under-18 Experience · FBI IC3, PSA I-060523-PSA · NCMEC, Take It Down.
Reused from this site’s detail pages, with the read dates recorded there: OpenAI, Terms of Use and Managing parental controls in ChatGPT · Meta, Meta AIs Terms of Service · Character.AI, Terms of Service · Anthropic, Age assurance on Claude · Federal Trade Commission, companion chatbot inquiry, all read 18 Sep 2026 · FTC, TAKE IT DOWN Act enforcement, read 22 Sep 2026 · Microsoft Research, VALL-E, read 10 Sep 2026.
The comparisons on this page (help desk and pen pal, film age label, garden fence, calculator and always-free friend, changing the locks, road safety) are this site’s own ways of explaining, not claims from these sources.
Free printable AI cheat sheets, and AI news every week.
Sixteen one-page AI guides and checklists in one free PDF, no sign-up. And if you want to keep up, the newsletter covers what changed in AI each week, in plain English, with the sources.
SUBSCRIBE FREE ↗ GET THE CHEAT SHEET PACK