Quick answer
Most major AI chatbots share a small set of privacy settings worth checking tonight: chat history (keep it or auto-delete it), training opt-out (whether your conversations are used to improve the underlying model), memory controls (what the AI remembers about your teenager across sessions), and account age (which most tools ask for but few verify). None of these take more than ten minutes to find and adjust, and none of them require technical knowledge. The value is not that any single setting is powerful. It is that stacked together, they meaningfully shrink what a chatbot retains about your teenager, which matters more the more they use it.
Why these settings matter more than they look
A conversation with an AI chatbot feels private in the moment - just your teenager and a screen - but it isn't private in the way a diary is. Depending on the settings, conversations can be stored, reviewed to improve the product, and remembered across sessions in ways your teenager may not notice happening. None of this is sinister by design. It is simply how these products are built, and privacy settings exist precisely because that default is not always what a family wants.
The stakes are higher for a teenager than an adult, for one simple reason: teenagers disclose more. A chat interface invites a conversational, unguarded tone that a search box never did, and what gets typed in that unguarded moment - a worry, a personal detail, a friend's name - is exactly the kind of information these settings are designed to contain.
None of this needs to feel alarming. It is closer to locking a front door than installing a security system: a small, sensible habit rather than a sign of danger. The point of running through the checklist is not that your teenager's chatbot use is unsafe by default. It's that the defaults are set by a product team optimising for engagement and product improvement, not for your family's privacy preferences, and a ten-minute check brings the settings back in line with what you'd actually choose.
The settings that exist across most major chatbots
| Setting | What it does | Why it's worth checking |
|---|---|---|
| Chat history | Stores or deletes past conversations | Old conversations can otherwise sit there indefinitely |
| Training opt-out | Excludes your conversations from being used to improve the AI model | Keeps what's typed from becoming training material |
| Memory controls | Lets the AI recall details across separate conversations | Worth knowing what's remembered, and clearing it periodically |
| Account age | Sets a birthdate at signup, which can switch on different safety defaults | Rarely verified, but still worth setting accurately |
| Data export or deletion | Lets you download or permanently remove account data | Useful when closing an account or starting fresh |
These settings differ in exact wording and location between tools, and they move around as products update, so treat the table as what to look for rather than a fixed map of menus.
The do-this-tonight checklist
- Open the settings or privacy menu in whatever chatbot your teenager actually uses - ask them to show you, which doubles as a useful conversation starter.
- Find the training opt-out and turn it on if it isn't already. This is usually the single highest-value setting in the list.
- Check chat history. Decide together whether to keep it (useful for continuity in schoolwork) or auto-delete it after a set period.
- Look at memory. Read what the AI has stored, delete anything that shouldn't be there, and agree how often to review it.
- Confirm the account age is set honestly. It won't summon a fortress of protection, but an accurate age is still the right default.
- Note the date you did this. Settings change with product updates, so a rough six-monthly recheck is worth scheduling.
What these settings don't do
Be honest with your teenager about the limits here, because overselling privacy settings creates false confidence. Turning off training opt-out does not make a conversation confidential in a legal sense. Deleting chat history does not guarantee nothing was retained before deletion. And no setting stops your teenager from sharing something they shouldn't in the moment - that is a judgement question, not a settings question, covered in full in parental controls for AI tools: what exists and what doesn't.
The deeper habit worth building alongside the checklist is treating an AI conversation the way your teenager would treat a comment posted somewhere semi-public: worth thinking about before typing, not after. That thinking extends naturally into what teens share without realising.
It is also worth checking settings on a shared or family device separately from a personal phone. Memory and history can behave differently depending on whether an account is logged in on a browser everyone uses, and a setting adjusted on one device does not automatically carry over to another.
Common mistakes parents make
- Never checking at all, on the assumption that defaults are fine. They usually are not the most private option available.
- Doing it once and forgetting it. Settings reset or change with product updates more often than most families expect.
- Treating settings as a substitute for the conversation about what's actually wise to share.
- Only checking the tool you know about. Teenagers often use more than one chatbot, each with its own settings.
- Assuming stated account age does anything meaningful for safety. It rarely does, on its own.
- Checking one device and assuming it's done. A phone, a school laptop and a shared family computer can each need their own pass.
The recommendation: set aside ten minutes tonight, sit with your teenager, and work through the checklist together on whatever AI tool they actually use. It is a small, concrete action that measurably reduces what gets stored and retained, and doing it together turns a privacy chore into the start of a much better ongoing conversation about AI, covered more broadly in AI education for teenagers in Australia.
Frequently asked questions
Related insights
Written by
Andrew Chisholm
Andrew Chisholm writes for Edison AI Insights on AI in education - how schools, teachers and students build genuine capability rather than quiet dependence.
Published by Edison AI Academy · About the academy
Learn AI the Edison way, with judgement built in.
Edison AI Academy teaches ambitious Australian students to think, build, and lead with AI through structured, project-based, responsible education.
