Family AI Rules: Agreeing How AI Is Used at Home and at School

Additional resource for the lesson “AI in the Family and at School” — Online Privacy course

AI chatbots often reach the kitchen table and the classroom before anyone has talked about them. Family AI rules do not need to be long or strict. This resource helps you agree a few together, check the age limits the services set, and know what your children’s school may ask of you, so that decisions about your family’s data are made on purpose.

A. Why this matters

Earlier resources in this section looked at where your AI conversations end up and what to keep out of them. In a family there is one more layer: the person typing may be a child, and what they type is often about the people they live with.

A homework piece about “my family” or a birthday message for a grandparent can carry an address, a school name or a relative’s name. This is not a reason for alarm. It is the question this course keeps asking: what is known about you, by whom, and whether anyone chose to share it.

Schools meet the same question from the other side: when AI tools are used in class, pupils’ work and data can be involved. Knowing the guidance helps you ask useful questions, calmly.

What the sources say

Internet Matters, a UK not-for-profit organisation that supports families with children’s online safety, published research on children and AI chatbots in 2025. It reports that “nearly half of children using AI chatbots say they use the tools for schoolwork”, and that two in five children who use them have no concerns about following the advice they receive.

The NSPCC, the UK children’s charity, suggests explaining to children that AI “can be useful but isn’t always accurate”. It also notes that “it can help to agree as a family on when it’s ok to use AI for schoolwork”, and encourages children not to share personal information.

For schools, the Department for Education in England recommends that personal data is not used in generative AI tools, and asks schools to comply with the age restrictions set by AI tools.

Behind this sits a principle. The UK’s Information Commissioner’s Office (ICO) has a statutory code of practice for online services likely to be accessed by children, the Children’s code, which “seeks to protect children within the digital world, not protect them from it.”

B. Key concepts

Five ideas for treating AI at home as a privacy question.

A family AI agreement

A short list of rules the household agrees together: which AI tools, for what, what stays out of the chat, and who to come to with a worry. It can fit on a note on the fridge.

Why it matters to you: A rule that has been talked through can be explained, and a rule that can be explained can be adjusted when the tools change.

Age limits in the terms

The minimum age a service sets in its terms of service, the rules you accept when you create an account. Limits differ by service and can depend on your country.

OpenAI’s terms say you must be at least 13, or the minimum age required in your country to consent, and that under-18s need a parent’s or legal guardian’s permission. Anthropic’s consumer terms set the minimum at 18, or the minimum age in your location if higher. Google’s help pages say you must be 13 (or the applicable age in your country) to use Gemini with a personal or school Google Account. Google also lets parents give children under 13 access through Family Link, its app for managing a child’s account, but says this is not available for supervised accounts in the UK, the European Economic Area and Switzerland.

Why it matters to you: An age limit is the service’s own statement of whose data it expects to handle: a fact to know before you decide, not a verdict on your family. And when a child uses a parent’s signed-in account, their conversations end up in that account’s history, mixed with the parent’s.

Family information

The details a child’s chat can carry about other people: where the family lives, which school, who is at home. Google’s own help page for parents suggests telling children “not to share sensitive personal information like their address, school name or family information”.

Why it matters to you: Privacy in a family is shared. One person’s chat can carry the details of people who did not use the tool and were not asked.

A service, not a person

Google’s guidance for parents puts it simply: “although Gemini is conversational, it’s not a person”. The NSPCC makes a similar point: AI tools are not trusted adults, so children should not rely on them for support.

Why it matters to you: Something that feels like a friend can invite confidences. Knowing that a chatbot is a service run by a company helps a child choose what to tell it.

School use and parents’ agreement

The Department for Education’s guidance asks schools to make sure pupils and parents understand when personal data is processed in AI tools, and to seek agreement. It also says schools must not allow pupils’ original work to be used to train generative AI without permission, which for a child can come from a parent or legal guardian.

Why it matters to you: You have a recognised role. Asking the school how it uses AI is part of the arrangement the guidance describes, not a complaint.

C. A practical example: Lucia and the homework chatbot

Lucia is 48, a parent who uses online services every day. One evening she notices that her children have been using a chatbot for homework, sometimes on her own phone, signed in to her account. She has no wish to forbid it, but she wants to know what is being shared, and with whom.

At a parents’ evening she mentions it to Sara, 28, a teacher at the school. Sara explains the school’s approach: staff are asked not to put pupils’ names or details into AI tools, and the school would ask before using pupils’ work with any AI service. She adds that pupils sometimes paste in whole messages, other people’s names included, when a summary would do.

What changed after that conversation

Lucia did the exercise below one Sunday, with her children at the table. It was a planning session, not a telling-off.

• She read the chatbot’s terms: under-18s needed a parent’s permission. She gave it knowingly, not by accident.
• She stopped lending her signed-in phone for homework, so their conversations and hers were no longer mixed.
• Together they wrote three family AI rules for the fridge: AI explains and checks, it does not do the whole task; names, the address, the school and family matters stay out; anything worrying goes to an adult.
• She emailed the school to ask which AI tools are used in class and whether pupils’ work is shared with them.
• She set a date to review the rules next term.

Sara began opening AI lessons with a short reminder of what to leave out of a prompt. Neither aimed for perfect control, only for choices they could explain.

D. Try it yourself: draft a family AI agreement (20 minutes)

Five short steps, best done with the people who will follow the rules. Nothing you find means anyone has done something wrong.

Step 1 — List the AI tools in your home (4 minutes)

• Write down the AI chatbots and assistants used at home, including those inside apps you already use.
• Next to each, note whose account it runs on and who uses it, including tools the school has asked for.

Step 2 — Find the age limit for each (5 minutes)

• Open each service’s terms of service or its help pages for parents, and look for the part about creating an account or age requirements.
• Write the age next to each tool, and whether it asks for a parent’s permission.
• Where a child is under the limit, or uses an adult’s account, decide together what to do.

Step 3 — Agree three rules (6 minutes)

• One rule about when: for example, AI can explain or check, but the work stays yours.
• One rule about what stays out: names, the address, the school, family matters, other people’s messages and photos.
• One rule about who to come to: a trusted adult, for anything worrying, personal or strange.
• Three rules you remember beat ten you do not.

Step 4 — Prepare your questions for the school (3 minutes)

• Which AI tools are used in class or for homework?
• Is any pupil’s personal data or original work entered into them, and will parents be asked first?
• How does the school check that tools are used within their age limits?

Step 5 — Set a review date (2 minutes)

• Choose a date, such as the start of next term, to look at the agreement again.
• Put it in a shared calendar, so that reviewing becomes routine, not a reaction to a problem.

The agreement is a starting point, made together.

E. Videos, articles and further resources

Guidance first, then children’s online safety organisations and the services’ own terms and help pages. All in English.

Education and data protection guidance

Department for Education (England) — Generative artificial intelligence (AI) in education
How schools should handle personal data, age restrictions and pupils’ work with generative AI.
https://www.gov.uk/government/publications/generative-artificial-intelligence-in-education/generative-artificial-intelligence-ai-in-education

Information Commissioner’s Office (UK) — Age appropriate design code: executive summary
The Children’s code in brief: 15 standards for online services likely to be accessed by children in the UK.
https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/childrens-information/childrens-code-guidance-and-resources/age-appropriate-design-a-code-of-practice-for-online-services/executive-summary/

Children’s online safety organisations

NSPCC — Talking to children about AI
Practical ways to talk about AI at home, including agreeing as a family when AI is fine for schoolwork.
https://www.nspcc.org.uk/keeping-children-safe/online-safety/talking-to-children-about-ai/

Internet Matters — Me, Myself and AI: chatbot research
UK research on how children use AI chatbots, with recommendations for parents and schools.
https://www.internetmatters.org/hub/research/me-myself-and-ai-chatbot-research/

Service terms and help pages

OpenAI — Terms of Use (EEA, Switzerland and UK)
The minimum age for its services and the parental permission required for under-18s.
https://openai.com/policies/eu-terms-of-use/

Anthropic — Consumer Terms of Service
The minimum age for using its consumer services.
https://www.anthropic.com/legal/consumer-terms

Gemini Apps Help — What you need to sign in to Gemini Apps
The age requirements for using Gemini with a personal, school or work Google Account.
https://support.google.com/gemini/answer/13278668?hl=en-GB

Gemini Apps Help — Guide your child’s Gemini Apps experience
Google’s guidance for parents: Family Link access, content filters and what children should not share.
https://support.google.com/gemini/answer/16109150?hl=en-GB

Links checked in October 2026. If an address changes, searching for the title on the organisation’s site usually finds it.

F. The Cyber Welfare Framework: Skills, Awareness, Secure Behaviour

This resource sits on the Awareness pillar at level FL2 – Beginner: understanding how AI tools touch a family’s data before setting rules around them.

Skills

• Finding the minimum age in a service’s terms or help pages.
• Writing a short family AI agreement the whole household can explain.
• Preparing calm questions for a school about its use of AI.

Awareness

• Recognising that a child’s chat can carry details about the whole family.
• Knowing that age limits are set by each service and can vary by country.
• Knowing that schools have guidance on pupils’ data and work, with a role for parents.

Secure Behaviour

• Keeping names, addresses, the school and family matters out of AI chats.
• Giving each user their own appropriate account rather than sharing an adult’s.
• Reviewing the agreement now and then, because tools and children change.

G. Questions to sit with

• In your household’s last few chats, which detail would you most want left out?
• Which of your family AI rules would your children say is fair, and which would they want to change?
• What would you want to know from the school before pupils’ work goes into an AI tool?

H. What to do now

The recommendations (R) and security measures (MS) from the Cyber Welfare database that apply most directly here.

1. The family agreement

No recommendation in the database covers this part: the work is the agreement itself. Make the “what stays out” rule specific enough for a child to apply without asking.

Minimum commitment: this week, agree one rule with your family about what stays out of AI chats.

2. The devices and apps your family shares

• R13 — Check what your apps can do: review the permissions of AI apps on shared phones and tablets, such as microphone, camera and photos, and lower those they do not need.
• R22 — Remove the apps you no longer use: if an AI app was tried once and forgotten, take it off the family device.

This resource looks at the privacy side. For questions about a child’s safety or wellbeing online, competent organisations can help: the NSPCC offers a helpline for adults worried about a child (0808 800 5000, 10am to 4pm, Monday to Friday) and Childline for children and young people (0800 1111); Internet Matters publishes guides for parents; and the ICO explains the Children’s code. Outside the UK, look for your country’s equivalent services.

Minimum commitment: on one shared device, check which AI apps are installed and whose account each one uses.

In short

• A child’s chat can carry details about the whole family: names, the address, the school.
• Each AI service sets its own minimum age, which can vary by country: check it before deciding.
• Guidance for schools in England covers pupils’ data, age limits and pupils’ work, and gives parents a role.
• Good family AI rules are few, agreed together and reviewed now and then.

Related resources in this course

• AI Chatbot Privacy: Where Your Conversations End Up
• What Not to Share with AI Chatbots: Keep, Generalise, Share
• Opt Out of AI Training: Check Your Chatbot Settings

Discover more companion resources from the online courses of the Protect Your Digital Privacy programme.