Families and schools
What are the risks of AI for children, and how can we prepare them?
What is the short answer?
Useful starting points are sharing personal information, accepting incorrect answers, encountering inappropriate content and placing too much trust in a tool. Preparing children means helping them pause, check and ask for help through age-appropriate activities and boundaries.
Where should the conversation start?
Start with a specific question: what do you want to use this tool for? Inventing a character with a teacher raises different questions from talking alone with a chatbot about a personal problem. Before choosing an app, agree on the task, the adult support and the information that stays out.
UNICEF's guidance on AI and children includes safety, privacy and preparing children among its requirements. This is policy guidance, not a certification that a particular app is safe.
Source: UNICEF — Guidance on AI and children, version 3.0 (2025)
What information should stay out of an AI conversation?
A learning activity can use fictional characters. A story does not need a student's full name, home address, daily schedule or class photograph. Practicing dialogue does not require uploading a private conversation.
- Replace real names with fictional characters.
- Review text and files before sharing them.
- Ask the responsible adult when a tool requests unexpected information.
How can children learn to question an answer?
Prepare a fictional paragraph containing a simple error, then ask the group to check it against a book or a source selected by the teacher. Compare the original paragraph with the correction. Practicing critical thinking does not require exposing children to dangerous content.
The useful question is: how could we check this? Friendly wording, a date and accurate spelling do not prove that an answer is true. A reference must also be opened and checked; its presence alone does not support the claim.
How should the activity change with age?
These are teaching suggestions, not permission to access a service. Accounts and tools must follow provider requirements and school rules. Students can practice these habits without individual accounts.
- In elementary school: guided demonstrations, visual examples and a simple rule—ask when something is confusing.
- With preteens: compare answers, spot personal information and explain what they would change.
- With teenagers: review sources, authorship and publishing decisions, including who could be affected.
What if something uncomfortable appears?
Agree on an exit before starting: stop the activity, close the content and tell a trusted adult. Make it clear that asking for help is the right response, even after a curious click. The adult chooses the next step under the school's or family's reporting procedure.
How does Defend Yourself Against AI put this into practice?
AI with Marcé uses guided situations to explore what information to share, how to question content and when to ask for help. Examples are adapted to the group's ages. The purpose is to build habits participants can explain and practice, without promising complete protection.
Defend Yourself Against AI. Explore the program and its PDF →Which sources did we consult?
How do we put these answers together?
We check factual claims against the linked sources. Activities are educational suggestions from AI with Marcé, not research findings or undocumented classroom experiences. Dates indicate publication and content changes. Email us to suggest a correction. Email Alex.