# AI Activities for Kids: 10 to Try Together This Weekend **Author:** Dan Fitzpatrick **Canonical URL:** https://parents.theaieducator.io/posts/ai-activities-for-kids **Published:** 2026-10-20T07:24:00.000Z **Updated:** 2026-10-20T07:19:04.551Z **Category:** AI for Parents ## TL;DR Yes, there are good things to do with AI and your child this weekend, as long as you hold the keyboard yourself. With a child under 13, type, read each answer before they do, and let them decide what to ask and whether it is right. Common Sense Media advises no AI chatbots for children 5 and under, so the ten ideas start at 6. Keep personal details out, check anything that matters, and stop if the chat turns toward feelings or friendship. Your school's rules, a platform's age limit and a doctor's or safeguarding lead's advice come first, and no setting makes a chatbot safe for a young child. ## Key takeaways - Hold the keyboard yourself: for a child under 13, the adult types, reads each answer first, and the child decides what to ask and whether it is right. - Common Sense Media (page reviewed June 8, 2026) advises no AI chatbots for ages 5 and under, and limited use alongside a parent or caregiver for ages 6 to 12. - Pick activities where your child does something before the machine speaks and something after it stops, such as guessing first, catching a mistake or teaching it back. - Keep names, school, address, photos and secrets out of every chat, and treat cooking times, allergies and opening hours as unchecked until you confirm them elsewhere. - Close the chat if it drifts toward feelings or friendship, and talk to your child yourself. ## Article Yes, there are good things to do with AI and your child this weekend, as long as you hold the keyboard. With a child under 13, I would type, read each answer before they do, and let them decide what to ask and whether it is right. Common Sense Media advises no AI chatbots for children 5 and under, so these ten ideas start at 6. ## What are the ground rules before we start? Four rules cover every activity on this page. Some of what sits behind them is evidence from a named source, and some is my own suggestion, so I have marked which is which. **What the sources say.** [OpenAI's help center](https://help.openai.com/en/articles/8313401-is-chatgpt-safe-for-all-ages) says "ChatGPT is not meant for children under 13," and adds that for children under 13 in an education setting, an adult must conduct the actual interaction with ChatGPT. I found no OpenAI statement about a parent typing at home while a child watches, so I treat the education rule as the standard to copy. [Google's family guidance for Gemini](https://support.google.com/families/answer/16109150?hl=en) (the page shows no date) says "Our filters try to limit inappropriate content from appearing, but they're not perfect." [Common Sense Media](https://www.commonsensemedia.org/articles/ai-chatbots-and-your-child-age-and-stage-guidance), on a page last reviewed June 8, 2026, recommends limited use alongside a parent or caregiver for ages 6 to 12, and no AI for emotional support or friendship for anyone under 18. A University of Cambridge paper in *Learning, Media and Technology* (2024), [as reported by ScienceDaily](https://sciencedaily.com/releases/2024/07/240710195430.htm) and built from documented cases rather than a survey, says that "children are much more likely than adults to treat chatbots as if they are human." **What I suggest.** I call the first rule the Keyboard Rule: you hold the keyboard, they hold the idea. For a child under 13, the adult types, reads each answer first and reads it aloud, and the child decides what to ask and whether the answer is right. The other three rules go with it. - **Nothing personal goes in.** No names, school, address, photos or secrets. Google's guidance tells parents to say the same to a child using Gemini. - **Check anything that matters.** AI is a guessing machine. It has read more words than any person ever could, and it uses all of them to guess what should come next. It is very good at guessing. It does not know when it is wrong, so checking is your job. The three questions I use, in order: Who says? How could we check? What would it look like if this were wrong? - **Stop when it turns into company.** If the chat drifts toward feelings, friendship or secrets, close it and talk to your child yourself. You are also on an adult account, which is another reason to read first. Nothing here tells me it has been tuned for a six year old. ## What can we actually do together this weekend? Each idea names the age it suits, a prompt you can copy, and the part that stays with your child. That last part is the point. My governing principle for families is: let AI help with the doing; keep the thinking with your child. ### 1. The bedtime story your child edits (ages 6 to 9) Let your child choose a hero and a problem. You type: *Write the first four sentences of a bedtime story about [hero], who [problem]. Stop there.* Read it out. Your child fixes one line they think is dull or wrong, then tells the rest of the story in their own voice, with the screen off. **The child's part:** the plot, the fix, and the ending. ### 2. Catch the bluff (ages 8 and up) Pick a subject your child knows better than you do: a game, an animal, a band. Type: *Tell me five facts about [topic].* Their job is to find the one that is wrong, using the three questions. If they find nothing, that is a result too. Choose a harder topic and go again. **The child's part:** being the expert in the room. ### 3. Guess first, then ask (ages 6 and up) Whenever your child asks "why", say "what's your guess?" before you type anything. Then ask the AI and compare. The Wondering Gap is the few seconds between a child asking a question and somebody answering it. That gap is where the thinking happens. A guess before the answer keeps it open. **The child's part:** the guess, and deciding who was closer. ### 4. Teach it back (ages 9 and up) Pick something from school this week. Type: *Explain [topic] to a 10 year old in five short sentences.* Then turn the screen over. Your child teaches it to you in their own words, and you ask one "why" question. **The child's part:** all of the explaining. If they cannot do it without the screen, you have found what needs another look. ### 5. The fridge-raid dinner (ages 7 and up) List what is in the kitchen and type: *We have [ingredients]. Suggest three simple dinners.* Your child picks one, helps shop for the gaps and does the cooking that is safe for their age, while you handle knives and heat. Treat any cooking time, temperature or allergy detail as unchecked until you have confirmed it from a recipe you trust. **The child's part:** the choice, the shopping list and the cooking. ### 6. The rainy-day planner (ages 9 and up) Type: *Give me ten ideas for a rainy Saturday indoors for children aged 7 and 11.* Leave your location out. Your child ranks the ideas, then checks the real opening hours and prices on real websites. The AI can suggest a place that has closed. **The child's part:** the ranking and the fact-checking. ### 7. Invent a family game (ages 6 and up) Type: *Suggest rules for a new card game for three players, using a normal deck, for ages 6 to 12.* Play it once. Your child changes the rule that did not work, and you ask the AI to rewrite the rules with their change. They decide whether the new version is better. **The child's part:** playtester and rule-maker. ### 8. Draw first, describe second (ages 6 to 10) Your child draws an animal that does not exist and describes it out loud. You type the description (not a photo of the drawing) and ask: *Write a short field-guide entry for this animal.* Your child corrects what the AI got wrong and adds one fact of their own. **The child's part:** the creature, and the corrections. ### 9. Argue the other side (ages 11 and up) Choose a family question with no right answer, such as whether phones should stay out of bedrooms. Type: *Give me the three strongest arguments against [position].* Your child writes or says a reply to each one. This is also a good way into the conversation about where your child already stands. **The child's part:** the rebuttal. ### 10. Quiz me, do not tell me (ages 13 and up) For a teen revising for a test, the prompt is: *Quiz me on [topic], one question at a time. Do not give me the answer until I have tried. After each answer, tell me what I missed.* They attempt every question before looking. Check anything surprising against class notes, because this is practice, not a source. Common Sense Media says easy AI answers can take away the struggle kids need to build skills, which is why the attempt has to come first. **The child's part:** every attempt, before the machine speaks. ## What I See in Schools The pattern I see most often when teachers try AI with a class is this. The lessons that go well put a child's own attempt, guess or knowledge on the table before the machine speaks. The lessons that go badly start with the machine's answer and ask the class to react to it. Most of the ten ideas above are built the first way on purpose. That is my observation from working with schools, not a research finding, and I would treat it as the reason for the design rather than proof that it works. > **A good AI activity at home is one where your child has done something before the machine speaks, and something after it stops.** ## How does this change by age? **5 and under.** Skip chatbots, as Common Sense Media advises. Reading aloud and storytelling do more at this age. A no-screen version of idea 3 works well: you play the guessing machine, finish your child's sentences with the most likely word, and let them catch you when you are wrong. **6 to 8.** Ideas 1, 3, 7 and 8, with you typing and reading everything first. Keep sessions short. **9 to 12.** Add ideas 2, 4, 5 and 6, and teach the three questions. [Internet Matters](https://www.internetmatters.org/hub/guidance/ai-chatbots-and-virtual-friends/) (its page carries no visible date) says that exploring AI chatbots with your child "can help you better understand the platform." Our [guide for 9 to 12 year olds](https://parents.theaieducator.io/posts/ai-for-9-to-12-year-olds) goes further. **13 to 15.** Ideas 9 and 10. OpenAI asks for a parent's permission for ages 13 to 18, so this is the stage to agree the rules out loud, and to ask your teen to show you how they use it. Our piece on [what teens actually use AI for](https://parents.theaieducator.io/posts/what-do-teens-use-ai-for) explains what you may find. **16 to 18.** They will run these themselves. Your part is to ask what they last used it for and what it got wrong. ## What could you say? Short, light, and not after a row. > **You could say:** "Want to try something? You tell me what to type, and we'll see if the computer gets it right. You're the boss." > **You could say:** "If it says anything wrong, you get a point. If it says anything weird or personal, you tell me and we close it." > **You could say (teens):** "Pick something you're revising. I won't read the answers. I just want to hear one thing it got wrong." > **You could say (if it asks about feelings or friendship):** "That's one for a person who knows you. Let's talk about it, you and me." ## What should we not do with AI at home? Do not use it as a friend, a confidant or a counselor, for any child. Common Sense Media recommends no AI for emotional support or friendship under 18, and our guide to [what to do when a teen talks to an AI about feelings](https://parents.theaieducator.io/posts/my-teen-talks-to-ai-about-feelings) covers the conversation that follows. If your child ever says, in words or in a chat, that they are in distress, talk to them first. In the US you can call or text 988. In the UK, Samaritans is on 116 123 and Childline on 0800 1111. Your doctor or your child's school safeguarding lead can help too, and a chatbot is not a clinician. Do not hand it the part of homework where the learning happens. The test I use is [the Homework Test](https://parents.theaieducator.io/posts/is-using-ai-for-homework-cheating): if the machine did this part, what would my child lose? If the answer is nothing, let it help. If the answer is the learning, that part stays with your child. And do not upload photos, voice recordings or schoolwork with your child's name on it. What happens to what you type is worth five minutes: our guide to [what ChatGPT collects about your child](https://parents.theaieducator.io/posts/what-data-does-chatgpt-collect-about-my-child) covers the two settings that change it. ## What none of this can guarantee No setting makes a chatbot safe for a young child. Answers can be wrong however confident they sound, and a parent's adult account has no child-specific protection that I can point to. Your school's rules come first: some teachers want any AI-assisted work disclosed, and a rule at school beats an idea on this page. The same goes for a platform's stated age limit, and for a doctor's or safeguarding lead's advice if you are worried about your child rather than about a setting. The age limit is a promise, not a lock, so the [Stated Age Check](https://parents.theaieducator.io/posts/can-kids-use-chatgpt) is worth two minutes tonight if your child already has an account. If your child's school has not yet told parents what AI use at home is welcome, or what it wants disclosed, that is part of the work I do with schools through [AI training for schools](https://theaieducator.io/ai-training-for-schools?utm_source=parents.theaieducator.io&utm_medium=referral&utm_campaign=ai-activities-for-kids). It is worth sending to your head of year or principal. I write a [Sunday newsletter](https://theaieducator.io/?utm_source=parents.theaieducator.io&utm_medium=referral&utm_campaign=ai-activities-for-kids#newsletter) that keeps teachers and parents up to date without the panic. Parents are very welcome. *Dan Fitzpatrick is a former secondary teacher, assistant headteacher and Director of Digital Strategy who now helps schools and families make sense of AI. [More about Dan](https://theaieducator.io/about?utm_source=parents.theaieducator.io&utm_medium=referral&utm_campaign=ai-activities-for-kids).* ## Frequently Asked Questions ### Is it safe to let my child use ChatGPT while I sit with them? It is safer than a child alone, but not safe in any guaranteed sense. OpenAI's help center says ChatGPT is not meant for children under 13, and says that for under-13s in an education setting an adult must conduct the actual interaction. I found no OpenAI statement about a parent typing at home while a child watches, so I suggest following the education rule: you type, you read each answer first, and you stop if anything is off. ### Should I use an AI made for kids instead of ChatGPT? Tools made for young learners may suit ages 6 to 10 better, and Internet Matters says to favor them for that age group, but I have not tested any, so I cannot vouch for one. The same rules apply to every tool: an adult in the room, no personal details, and answers checked before they are trusted. ### What if my child asks the AI to be their friend or tells it a secret? Close the chat and talk to your child yourself, without making them feel in trouble. Common Sense Media recommends no AI for emotional support or friendship for anyone under 18. If your child is in distress, in the US you can call or text 988, and in the UK Samaritans is on 116 123 and Childline on 0800 1111. Your doctor or the school safeguarding lead can help too. ### Does using AI at home break my child's school rules? It can, depending on the school and the task. Some teachers want any AI-assisted work disclosed, and some set tasks where AI help is not allowed at all. Ask the teacher, or the head of year in a UK school, what the rule is for homework, and follow the school's rule over anything on this page. ### What can I do instead of chatbots with a child aged 5 or under? Read aloud, tell stories, and play the guessing-machine game: you finish your child's sentence with the most likely word and they catch you when you are wrong. Common Sense Media advises no AI chatbots for children 5 and under, and these no-screen versions teach the same idea. ### How do I stop my child using AI to do their homework for them? Use the Homework Test: if the machine did this part, what would my child lose? If the answer is nothing, let it help. If the answer is the learning, that part stays with your child. The quiz and teach-it-back ideas on this page are built to put the child's attempt first.