AI & safety
PaperTim was built on one rule: the AI works for you, never around you. Tim talks only to parents, knows only what you choose to write, and nothing he drafts reaches paper without your approval. Here is exactly how that works.
How PaperTim keeps the AI in its place
No marketing gloss here. This page describes what the software actually does. Every claim below is enforced in code and covered by our release tests.
Only parents talk to Tim
There are no child accounts in PaperTim: none, not optional, not supervised. Your children never see the dashboard, never see the chat, never see a screen at all. Tim's conversations happen with you: he drafts plans and worksheets, and every draft arrives in blue and waits for your decision.
What Tim is told
When Tim drafts for a child, he is told exactly what you wrote on their profile: a first name or pseudonym, an age, a working level, an output language, their interests, anything you've flagged to avoid, and your own notes. That is the whole list.
What Tim is never told
No birthdates: Tim gets an age as a number, never the date it comes from. No surnames. No photos: the pictures you take for grading stay in your account for your eyes and are never sent to the AI. No sibling details ride along in another child's request, and nothing about your account or billing ever reaches a model.
Prefer not to use a real name? Use a pseudonym
The profile's name field accepts anything your child answers to: a nickname, a middle name, "Captain R". Worksheets print with it, Tim drafts with it, and you can change it whenever you like in the child's profile and privacy settings. Your child's real name never has to enter PaperTim at all.
You approve everything
The AI in PaperTim can only draft. It cannot publish a plan, print a page, change a grade or touch your family's records: those buttons are yours alone. Edit any draft, regenerate it, or throw it away; nothing counts until you approve it.
Who provides the AI
Text is drafted by Anthropic's Claude models; worksheet illustrations by Google's Gemini. Requests carry only the profile facts described above and the task you asked for. Want your family's data removed entirely? Write to help@papertim.com and we will handle it.
Judge Tim by his paperwork
The clearest safety story is using it: draft a week, read every word before it prints, and see who is in charge.
Questions, answered
Everything parents ask before their first week with Tim. Something else on your mind? Just ask.
Can my child chat with the AI?
No. There is no child-facing login, app or chat anywhere in PaperTim. The product's design keeps children on paper and parents at the controls. It is not a setting that can be switched off, because the capability does not exist.
What personal data does the AI receive?
Only the child profile you wrote: first name or pseudonym, age, working level, output language, interests, dislikes and notes, plus the task you asked for. Never a birthdate, surname, photo, sibling detail or anything about your account.
Are my grading photos sent to the AI?
No. Photos of finished work exist for your grading screen only. They are stored privately in your account and are never part of any AI request.
Do I have to use my child's real name?
No. Use any pseudonym your child answers to: it works everywhere a name appears, and you can change it at any time.
Does the AI decide what my child learns?
No. You choose the curriculum and activate its objectives; Tim drafts within what you activated and suggests, never decides. Every plan and every worksheet waits for your approval.
How do I get our data deleted?
Email help@papertim.com from your account address and we will remove your family's data. The photos, profiles and records in PaperTim are yours: leaving with them deleted is part of the deal.