Keeping children safe with AI apps: a guide for parents

Where children meet AI

• Chatbots used for homework, questions and conversation
• AI features inside apps they already use, such as social media, games and messaging apps
• Image and video apps that change faces, create pictures or add filters
• Companion apps: chatbots that act as a friend or partner. They remember past conversations and are built to keep the user talking

Many of these tools are useful. Some carry real risks for children, especially younger ones.

The main risks

• Wrong or harmful answers. Chatbots can give confident but false information, including on health. See hallucination.
• Unsuitable content. Safety filters are not perfect. Children can come across violent, sexual or frightening material.
• Emotional attachment. A companion app can feel like a real friend. Some children may share secrets with it or spend hours a day talking to it.
• Privacy. Children may type their name, school or address into an app without thinking about where it goes. They may also upload photos.
• Fake images. AI apps can create false pictures of real people from ordinary photos. Such pictures are used for bullying and blackmail.
• Strangers and scams. Some accounts and apps that call themselves "AI" are fronts for people trying to contact or cheat children.

Age limits

Most AI apps set a minimum age in their terms of use. Some are for adults only. Read the app's terms before your child signs up. If your child is under the stated age, the app was not designed for them. Signing up with a false age also removes the protections that do exist.

Check whether the app offers a family or teen mode. Phone and app store settings can limit which apps a child can download.

Protecting privacy

• Teach children never to share these with any chatbot or app: full name, school, home address, phone number, B-form details and photos of themselves.
• Turn off chat history or memory where possible. See AI and your privacy.
• Keep devices in shared rooms at home for younger children.
• Try an app yourself first, so you know what it does.

Talking to your children

Rules work better alongside conversation. Bans alone often push use out of sight.

1. Ask, do not accuse. "Which apps do you use? Can you show me?" works better than a lecture.
2. Explain that AI is not a person. It does not care about them, even when it sounds kind. It can also be wrong.
3. Agree simple rules together: which apps, for how long and what never to share.
4. Make it safe to tell you. Say clearly that they will not be in trouble for telling you about something upsetting online.
5. Keep talking. Apps change quickly. Check in regularly.

Warning signs

• Secrecy about a particular app
• Hiding the screen when you walk in
• Hours spent talking to a chatbot, especially late at night
• Talking about an AI as a close friend or partner
• Sudden fear, anxiety or withdrawal after using their phone
• Requests for money, top-ups or mobile wallet transfers they cannot explain
• Mention of pictures of them that they did not take

If something goes wrong

1. Stay calm. Your reaction decides whether they come to you next time.
2. Keep evidence: screenshots, usernames, links and dates. Do not delete them.
3. Report the account in the app and block it.
4. If your child is being threatened, blackmailed or shown fake images of themselves, report the crime: call the National Cyber Crime Investigation Agency (NCCIA) cybercrime helpline on 1799.
5. Get support. If your child seems distressed, speak to a doctor, school counsellor or other qualified professional. AI does not replace them.

Never pay a blackmailer. Paying rarely ends the demands.

For help with school work and AI, see AI homework tools.