AI chatbots and AI companions: what parents and educators should know

Homework helper or companion: not the same thing

"AI chatbot" covers a lot. Some are general assistants kids use to explain a math problem, brainstorm a story or check a fact. Others are companion apps, designed to play a character, remember you and chat like a friend or partner. Many social apps now have a chatbot built in.

The risks differ. A homework helper can be wrong or make things up. A companion is built to keep you talking, and some feel very personal. Common Sense Media recommends that no one under 18 use AI companions until stronger safeguards are in place. So a good first question is simply: "Which ones do you use, and what for?"

What teens report

In a 2025 Common Sense Media survey of 1,060 American teens aged 13 to 17, 72% had used an AI companion at least once and 52% used one at least a few times a month. That's a US survey, but the same apps are used in Canada.

The same survey offers some reassurance and some things to watch:

• Among teens who use companions, 80% said they spend more time with real friends than with AI.
• About a third of users had talked with a companion about important or serious matters instead of with a real person.
• About a third had felt uncomfortable with something a companion said or did.
• About a quarter had shared personal information with one.

In other words, for most teens AI is one more thing they try, not a replacement for people. The goal is to keep it that way.

What a chatbot is, and what it isn't

A chatbot is software that predicts which words should come next, based on huge amounts of text. It can sound warm, funny and understanding. It doesn't know your child, doesn't have feelings and can't take responsibility for its advice. It can also be confidently wrong.

MediaSmarts suggests helping kids see chatbots as toys or tools, not as friends or therapists, and not as a substitute for real friendships or qualified mental health support. With younger kids, sit with them the first few times they use a chatbot or voice assistant, and stay involved as they get comfortable. At any age, encourage them to check what a bot tells them against trustworthy sources.

Privacy: what gets shared and stored

Conversations with a chatbot can feel private, but they're shared with a company. Depending on the service, what's typed may be stored, reviewed or used to improve the product.

A simple house rule helps: don't tell a bot anything you wouldn't post publicly. That includes full names, school, address, photos and other people's private details. MediaSmarts suggests looking at a tool's terms of service and privacy policy together, even just the summary.

Some apps offer parent settings. On Snapchat, for example, Family Center lets parents restrict My AI, the app's chatbot, from responding to their teen's chats. Menus can move between versions, so check the app's own help pages for current settings.

In the classroom

Students are already using these tools, so naming them openly helps. A few approaches that fit most grades:

• Explain the difference between a general assistant and a companion app, and why a bot designed to keep you chatting deserves extra caution.
• Practise checking: have students compare a chatbot's answer with a reliable source and spot what it got wrong.
• Talk about what not to share, using the same "would you post it publicly?" test.
• Make clear that a bot is not a counsellor. Point students to people: a trusted adult at school, and services like Kids Help Phone.

If a student mentions relying on a bot for emotional support, treat it like any other sign they may need someone to talk to, and follow your school's usual support steps.

Signs to watch, and where real support is

Using a chatbot is not a problem in itself. Worth a gentle conversation:

• pulling away from friends or family to spend time with a companion
• getting upset or secretive when they can't access it
• treating a bot's advice on health, relationships or safety as the final word
• mentioning that a bot said something sexual, frightening or cruel

If you notice these, stay curious rather than alarmed. Ask what they like about it and what it gives them. Often the answer points to something real they need more of, like connection or someone who listens. MediaSmarts' advice is to make sure kids know they can tell you if a chatbot says something inappropriate or upsetting.

What Canada is doing

In June 2026, the federal government introduced Bill C-34, the Safe Social Media Act. Among other things, it would require certain AI chatbot services to reduce the risk of their chatbot communicating harmful content, and it would create a Digital Safety Commission to oversee the rules.

It is a bill, not yet law, and its details may change as it moves through Parliament. Whatever happens, the most important safeguard is still the conversation at home.