How TeenCue works

Most parents arrive here with a specific version of the same question: my teenager is using AI constantly, and I have no idea what for. This page is the whole mechanism: what TeenCue sees, what it deliberately does not, and where it stops being the right tool.

What TeenCue does

There are two pieces. A browser extension runs in Chrome, Firefox, Edge or Brave on the computer your child uses, and a dashboard in your parent account shows what it found.

When your child sends a prompt to one of the supported AI services, the extension recognises it and the prompt is classified. What comes back and gets saved is a description of the chat: which service, a category such as schoolwork or personal advice, a more specific topic, and a concern rating if something looked worth your attention. That description is what appears on your dashboard. The prompt itself is not kept.

Alongside the dashboard there is a Weekly Teen Brief: a summary of the week, read when it suits you. You also get a heads-up by email, and on your phone if you use the TeenCue app, a few minutes after a chat on a topic you chose is rated Significant, and after anything that suggests self-harm or that someone your child knows may be in danger. A heads-up says the service, category, topic and concern level, never what was typed. Between 9 PM and 7 AM a topic heads-up waits until morning on your phone; a self-harm heads-up does not wait. You can turn topic heads-ups off in your settings. Self-harm heads-ups stay on. TeenCue is not a crisis service.

How setup works

Setup takes about 5 minutes and happens once per computer. It needs physical access to the machine your child uses. There is no way to install TeenCue onto someone else’s computer remotely, which is deliberate rather than a gap.

1. Create a parent account and add your child. This makes them a private profile.
2. On your child’s page, create a setup link and send it to the computer they use for AI.
3. Open the link there in Chrome, Firefox, Edge or Brave and choose Add to Chrome (or Add to Firefox, Add to Edge or Add to Brave). Setup finishes by itself. No code to type.
4. That is it. Activity starts appearing the next time a supported service is used.

Which AI services it covers

TeenCue supports nine services: ChatGPT, Gemini, Claude, Microsoft Copilot, Character.AI, Grok, DeepSeek, Perplexity and Poe. Support for the Meta AI website is being verified.

If your child uses an AI tool that is not on that list, TeenCue will not see it. New AI apps appear constantly and the honest position is that this list is the coverage, not a claim to cover everything. AI apps parents should know about is a wider survey of what is out there.

Browsers and devices it covers

TeenCue works in Chrome, Firefox, Edge and Brave on a computer, a desktop or a laptop. It does not cover phones or tablets, it does not run in other browsers such as Safari or Firefox, and it sees nothing outside the supported AI services. The TeenCue iPhone app is coming soon.

That limit matters more than any feature on this page. A teenager who uses AI mostly on their phone is largely outside what TeenCue can see, and if that describes your household you should weigh a device-level tool first. How to monitor teen AI use compares the realistic options side by side, TeenCue included.

What TeenCue records

For each AI chat, one row is saved. It holds:

• The service: ChatGPT, Gemini, Claude and so on.
• A category, from a fixed set: schoolwork, coding, research, creative, general questions, entertainment, personal advice, relationships, companionship or other.
• A topic, more specific than the category.
• A concern type and a severity from 0 to 3: Normal, Low, Moderate or Significant.
• A confidence score for the classification, and when it happened.

That is the entire row. There is no field for anything else, which is the part worth understanding before the next section: the limits below are not settings, they are the shape of the database.

What TeenCue does not record

• Not the conversation. The prompt text is sent from the extension to our backend and on to Anthropic, which returns the classification. It is not stored and it is not logged, and there is no column in the database that could hold it.
• Not the AI’s replies. No copy of the chat, no archive, nothing to scroll back through.
• Not general browsing. No browsing history, no page titles, no URLs from anywhere other than the supported AI services.
• Not keystrokes, screenshots, audio, video, location or anything from other apps.

None of this is a plan tier or a toggle. If what you actually want is to read your child’s messages, TeenCue cannot do it and no upgrade will. We would rather you knew that now than after paying. The privacy policy sets out the same thing in formal terms, including who processes what.

Your child will know

TeenCue is not built for covert monitoring. The extension is visible in your child’s browser, setup requires their computer, and the product assumes throughout that they know it is there.

That is a position, not an unfinished feature. Monitoring that is discovered tends to cost more in trust than it ever returned in information, and a tool your child knows about is one the two of you can actually discuss. Should parents monitor teen AI use? works through that trade-off properly, including the case for not monitoring at all.

How the safety signals work

Every chat is rated for concern as well as topic, from 0 to 3. Anything at Moderate or above is grouped as needing attention on your dashboard, with the concern type, the service and the time.

A rating is a reason to start a conversation, not a conclusion about your child. The classifier is wrong in both directions sometimes: it misses things and it flags homework about a difficult novel. Treat it as a prompt to ask, and how to talk to your child about AI has openings that do not start an argument.

How it fits with Bark, Qustodio and Screen Time

TeenCue is narrow on purpose and does not replace a parental control suite. Tools like Bark, Qustodio, Screen Time and Family Link cover a far wider surface (time limits, app blocking, web filtering, messages, location), and if your main concern is overall device use, one of those should be your primary tool, not this.

What those tools generally cannot tell you is what an AI chat was about. A broad monitor reports three hours in a browser; it does not separate a history essay from something you would want to know about. TeenCue fills that one gap and expects the rest of your setup to stay exactly where it is.

The platforms’ own controls are free and worth turning on alongside it: ChatGPT, Gemini and Claude, which has none because it is 18+.