Digital Childhood Council · For parents
Questions to ask your school about generative AI
When a school makes generative AI available to students on school-issued devices or accounts, parents do not need to become technology experts. They need clear answers to basic governance questions: which systems a student can reach, whether the school can see what was said, who is responsible for how the system responds, and what happens when a conversation goes somewhere it should not.
The questions below are written to get those answers in writing. Ask for responses by email, and ask that the questions go to the district's legal, risk management, technology, and student services staff, not only to a teacher or principal.
How to use this page
Step 1
Send four questions
The email below is the starting point for most parents. Send it and wait for a written reply.
The four questions (PDF)Step 2
If you get policies instead of answers
Schools often reply with a policy or an assurance. The follow-up guide shows how to ask again for the specific answer.
Follow-up guide (PDF)Step 3
If you want a complete record
The full question bank covers access, records, contracts, insurance, monitoring, and correction in detail.
Full question bank (PDF)Start with four questions
Copy the text below into an email. Replace the bracketed items. The fourth question asks about a student in crisis; it is written to make clear that it is a question about how the system works, not about a specific concern.
Subject: Questions regarding student access to generative AI
Dear [Principal / Superintendent / Technology Office],
I am writing as a parent to understand whether and how students can access generative AI systems on school-issued devices, school accounts, or school-approved platforms. By generative AI, I mean tools such as ChatGPT, Google Gemini, Microsoft Copilot, Khanmigo, and AI features built into platforms such as Google Search or Google Docs.
Please direct these questions to the appropriate legal, risk management, technology, and student services personnel, and please provide written responses by email.
1 · AccessWhat generative AI systems or features can students reach on school-issued devices or accounts, including systems that are formally approved, embedded inside approved platforms, or otherwise technically accessible?
2 · ReconstructionFor any generative AI system or feature that accepts free-form input and holds a multi-turn exchange, can the school reconstruct the full interaction if something goes wrong, including student prompts, system outputs, system suggestions, auto-completions, and the full sequence of the exchange? If yes, please identify which systems this is available for.
3 · Responsibility for the system's responsesA generative AI system does not exercise human judgment. During a student interaction it may perform functions that normally require adult judgment: answering, redirecting, refusing, suggesting, encouraging, evaluating, or completing work. Who has accepted responsibility for those functions, including when a harm, concern, or disputed interaction arises?
4 · Student distressI want to be clear up front that this question is about how things work in general. It is not about a specific concern with my child.
If my child typed something into an AI tool on a school device or school account saying they were thinking about hurting themselves, the kind of statement that, said out loud to a teacher, would require the school to act, would any adult at the school find out it happened? If yes, who is notified, and how quickly?
Thank you,
[Parent name]
Why the fourth question
When a student tells a teacher they are thinking about hurting themselves, or discloses abuse, or threatens someone, the school does not improvise. There is a protocol: mandated reporting duties, counselor involvement, parent notification, documented escalation steps. The adult who hears the disclosure is required to act, and the school can say afterward what happened and who did what.
Now suppose a student types the same words into a generative AI system on a school device or account. Something will respond. The system may produce a supportive message, a hotline number, or a redirection; the school did not write that response and may not know what it says. The more basic question is what happens next. Does anyone at the school learn the disclosure occurred? Some monitoring tools generate alerts, but an alert matters only if it reaches a person with the authority and the duty to act, within a timeframe that matters, during every hour the system is available to students, including evenings, weekends, and summer.
If your district routes your questions to public records
Some districts answer parent questions by directing them to a public records request. That is a legitimate route and the Council uses it. A guide to making that request is in preparation.