Privacy notice
This explains what the beta records, what leaves our servers, and how long anything is kept.
Last updated 19 July 2026
Beta
This is a working beta shown for evaluation. It is not a live service, it has not been reviewed by a lawyer, and it must not be used by real children. The text below describes exactly how the beta behaves today so that nothing here is a promise we do not keep.
What a parent gives us
When you create an account we store your name, your e-mail address, and a one-way hash of your password. We never store the password itself.
For each child you add, you choose a display name, an age band, a reading level, the subjects that are allowed, a tone for stories, and a daily message limit. You may also set a Profile PIN, which is stored only as a one-way hash.
What a child’s use records
Messages sent and replies received are stored so the conversation can continue and so you can review it.
When our safety checks act on a message, we record only what happened — the kind of event, the verdict, and the time. We deliberately do not store the child’s words in that record.
We count how many messages a child sends each day, to enforce the limit you set.
What we send to OpenAI
This beta uses OpenAI to check messages for safety and to write replies. Every message a child sends, along with a short window of the recent conversation, is transmitted to OpenAI for that purpose.
If a child attaches a photo, that photo is transmitted to OpenAI to be checked and answered. We do not write the photo to disk; only a note that a photo was attached is kept in the transcript.
When a reply is read aloud, the text of that reply is transmitted to OpenAI to generate the audio. The audio is streamed to the browser and is not stored.
OpenAI processes this data under its own terms. A production service would require a separate agreement and a separate, explicit parental consent for this transfer; this beta has neither.
What we do not do
We do not show advertising, we do not sell or share personal information for advertising, and we do not build marketing profiles of children.
Our operators can see that a safety event occurred and on which account, but they cannot read a child’s messages and cannot sign in as a child.
How long we keep it
Chat messages are deleted automatically after 90 days.
Profile settings and usage counts remain while the account exists. Deleting a child removes that child’s conversations and stored messages with them.
Questions
Because this is an evaluation beta, there is no support desk yet. Direct any question to whoever demonstrated it to you.