Most letters home about AI answer a question parents are not asking. They list the tools, promise the data is safe, and say the school is excited about personalized learning. The parent reading one at the kitchen table wants something else: when my child sits down to do their homework tonight, what are they allowed to hand over, and what do they still have to do themselves?
So here is the short answer to how school leaders should talk to parents about AI. Stop sending an inventory, and send a letter that answers three questions in this order: what AI is doing in their child's day, what their child will still do without it, and where parents come in. The first is the one schools write. The second is the one parents came for. The third is the one that decides whether they believe you.
The templates going around this term will hand you the first. The other two are blanks only your school can fill, and filling them is what the rest of this is about: what parents are actually worried about, the six lines a letter home needs, a test for the message you last sent, and a worked example to hold yours against.
Why are parents so uneasy about AI in school right now?
Parents are uneasy because most of them have heard almost nothing from the one source they trust most, and this month the gap is being filled from elsewhere.
The trust is not in doubt. A Child Trends survey of 1,025 US parents (fielded December 2025, published July 28, 2026) found that 43% named school communications as their most trusted source of information about their child's education, "more than double any other category." Only small shares picked social media, news or AI tools. The researchers' summary is blunt: parents "are most likely to rely on and trust information about their child's education when it comes directly from their school."
Yet schools are mostly quiet on AI. In a National Parents Union survey of 1,511 US public school parents, conducted by Echelon Insights in February 2026, 47% said their school had not given them information about its AI policy, against 37% who said it had. Fifty-seven percent said they had never been asked for input on AI tools in school. Meanwhile 56% said their child uses AI chatbots.
The worry underneath that silence is specific, and the people writing the letters consistently underrate it. RAND's survey panels, published on September 30, 2025, found that 61% of parents were concerned expanded AI use would harm students' critical-thinking skills, against 22% of district leaders who shared that worry. A gap that wide between what families fear and what the senders fear explains most of the letters that land badly. They answer the sender's question.
Into that silence, September has poured two loud messages. On September 2, New York City announced a one-year moratorium on student-facing generative AI covering nearly 600,000 students, with limited pilots in high schools. On September 9, the Institute for Family Studies, whose authors make the case for a pause, published a YouGov poll of 2,000 parents, fielded in July, in which 75% said they would support a policy to pause the introduction of AI into schools. The full question wording is not yet public (the report is due in November), so treat that figure as a reading of mood rather than a verdict on any particular policy.
My interpretation: a pause is what parents ask for when they cannot see what their own school has decided. The same week, an IBM and Morning Consult survey reported by Education Week (September 9, 2026) found that 83% of educators feel confident teaching about AI, while 66% of parents believe educators can. I read that 17-point gap as a visibility problem more than a competence problem. Parents cannot trust work they have never been shown.
What do parents actually need to hear from school leaders?
Parents need three answers, and the one most letters leave out is the one they want most.
1. What AI is doing in your child's day. Not a product list, but scenes a parent can picture. For example: "Your child's science teacher uses an AI tool to write practice questions at three levels of difficulty. Your child never logs in to it." Or: "In ninth-grade English, students use a school-approved chatbot to argue against their own essay's thesis, in class, with the teacher reading the exchange." Separate the tools staff use from the tools children touch, because parents worry about the second kind and are rarely told the difference.
2. What your child will still do without it. This is the missing answer. Parents' deepest worry is not data protection, though that matters. It is that their child will stop learning to think. A school that can say "every piece of extended writing in grades six to eight is drafted by hand, in class, before any tool is allowed near it" has answered that worry with a fact. A school that cannot say anything like it has not yet made the decision that parents are asking about.
The question behind this list is one I have asked school leaders before, as The Handover Question: "The Handover Question is the one test I ask school leaders to put to every task children are set: if a machine did this for them, what would they lose? If the answer is nothing, hand it over and use the time better. If the answer is the learning itself, protect it, and be able to say why." The parent letter is where "be able to say why" gets tested.
3. Where you come in. A named person to contact with a concern, not a general inbox. What happens if something goes wrong, and who answers for it, which is worth settling before a parent is the one who asks. And, most often missing, when the next decision about children's use of AI will be made and how a parent can have a say before it is. A report published in August 2026 by the nonprofit Bellwether, Agency in the Algorithm, makes the case that families should be engaged "early, before decisions are finalized," and argues that "the stronger path is to equip parents with the information and agency they need to weigh in effectively."
Does your message to parents pass the Kitchen Table Test?
The simplest check on anything a school sends home about AI is whether a parent could use it to answer their own child's question.
I suggest leaders think of it as the Kitchen Table Test. The Kitchen Table Test is the question I ask school leaders to put to anything they send home about AI: could a parent, reading it at the kitchen table, answer their own child's question, "Am I allowed to use AI for this?" If they could, the school has equipped a family. If they could not, the school has only informed parents about itself, and they will fill the gap from the news.
It is the family's version of a test I use with staff. The Monday Test is a single question to ask of any AI strategy before it is signed off: could a member of staff read it and know what to do differently on Monday morning? A strategy that fails the Monday Test usually fails the Kitchen Table Test too, and for the same reason: nobody has decided what the rules actually are.
Here is the distinction I would put on the wall of any leadership team meeting about this. A letter that lists tools tells parents about the school. A letter that passes the Kitchen Table Test helps them parent.
What should a school's letter to parents about AI actually say?
Six lines, in plain words. A letter with all six can be acted on by a family that same evening. A letter missing one of them tells you, fairly reliably, which email you are about to receive.
- Which AI staff use, and whether it goes near your child's data. Name the job it does, not the vendor's promise.
- Which AI children use, where, and who is in the room. Classroom, supervised, homework, never: say which.
- What children will still do unaided, and why. One protected task, named, beats a page of principles.
- The homework rule, in a sentence a twelve-year-old can repeat. This is the line parents will actually quote at the kitchen table.
- A named person, not an inbox. With a role, so a parent knows the seniority of the answer they are getting.
- The date of the next decision, and how to have a say before it. A date converts a worried parent into a participant.
The best free template I have found is TeachAI's sample letter to parents and guardians, part of its AI Guidance for Schools Toolkit. It is a good document and it is candid about its own limits. It sets out seven principles, notes that ChatGPT currently requires users to be at least 13, and then leaves blanks for the tools you actually use, your own responsible use policy and your signature. Those blanks are the work. A template can write lines one and two for you. Lines three to six are decisions, and nobody outside your school can make them.
A letter that passes is therefore shorter than most, specific enough to act on, and honest about what has not been decided.
Here is a hypothetical example for a middle school, written to show the shape rather than to be copied.
The typical version: "We are excited to be using a range of innovative AI tools to personalize learning. All tools have been carefully vetted for data privacy and safety. Please contact the school office with any questions."
A version that passes: "This year, teachers will use AI to prepare some materials, such as practice questions and reading passages at different levels. In grades seven and eight, students will use one school-approved chatbot in class, with a teacher present, to test their arguments in history and English. They will not use it for homework. All extended writing will still be drafted by hand in class first, because we want every child to learn to build an argument alone before a machine joins in. If your child asks whether they can use AI for a piece of homework, the answer is no unless their teacher has said yes in writing. Ms. Rivera, our assistant principal, is the person to contact with any concern. In January we will decide whether to extend chatbot use to grade six, and before we do, we will hold an evening where you can see the tool and tell us what you think."
The second version names a use, a limit, a reason, a rule a child can follow, a person and a date. A parent can repeat it to a twelve-year-old.
What I See in Practice
Across the leadership teams I work with, the parent letter is usually the last thing written, drafted by whoever has a free afternoon after the decisions are made. It inherits the language of the vendor materials sitting on the desk, which is how "innovative," "personalized" and "vetted" end up in so many of them.
The schools that find this conversation easiest have almost always made one decision first: what their children will not hand over. Once that list exists, the letter nearly writes itself, because the school has something to say that a parent cannot read anywhere else. Where that list does not exist, no amount of careful wording rescues the letter, because the gap is in the decision, not the prose.
I have written about AI and schools for Forbes and in my books for several years, and the public conversation shows the same pattern from the outside. The loudest arguments are about tools and bans. The questions parents ask underneath them are about children: will mine still learn to write, to reason, to be bored and push through it? Schools that answer the underneath question rarely have to fight the loud one.
Should you invite parents into AI decisions when so many want a pause?
Yes, because the alternative is not a calmer conversation. It is the same conversation, held without you.
The counterargument deserves a fair hearing. If three in four parents in one poll say they would back a pause, opening the door looks like inviting a veto. Leaders also know that consultation can stall a decision for a year.
Two things answer it. First, consulting is not the same as holding a vote. Leaders still decide, and they still have to be able to explain the decision. What changes is that parents hear the reasoning before the decision, when it can still use what they know about their own children. Second, not every decision needs the same level of consultation, a point the Bellwether report makes as "match engagement to stakes." Here is how I would draw the lines:
| Decision | What parents need |
|---|---|
| Staff use AI to prepare materials; no student data involved | Told, clearly, once a year |
| Students use an approved tool in class, supervised | Told before it starts, with a demonstration |
| AI does work students would otherwise do, or holds their data | Asked before the decision, with a date for the answer |
Consultation after the decision is not consultation. It is an announcement with a reply address, and parents can tell the difference.
What should school leaders do this term?
Start with what you have already sent, and work backward to the decisions it exposes.
- Test your last message. Read whatever your school last sent home about AI, count how many of the six lines it contains, and put the Kitchen Table Test to it. If a parent could not answer their child's question from it, rewrite it.
- Write the list of what children will not hand over. Do it with teachers, subject by subject, using The Handover Question. It is the answer parents most want and the one you cannot borrow from anyone else.
- Name a person. One named contact for AI concerns, published on the website and in the letter.
- Put the next decision on the calendar, and the parents in front of it. Choose the next change to children's use of AI and set an evening, before the decision, where families can see the tool working and say what they think.
- Publish what is approved. If you have an approved list, share it with families. If you do not, what schools should require before approving an AI tool sets out how to build one.
If restriction is on the table, the parent conversation belongs inside that decision too. Should schools ban AI? sets out the three answers any restriction needs, and families should be able to hear all three. For the bigger question under the protected list, see what schools should be preparing children for now. And if the letter keeps failing because the strategy behind it is vague, begin with how to create an AI strategy that people actually use.
Where to go from here
Parents are not asking schools to understand AI better than anyone else. They are asking to be told what their own school has decided, and to be asked before the next decision is made.
If your leadership team is working out what to say to families about AI, and what it needs to decide first so there is something worth saying, this is part of the AI strategy work I do with schools. For more on the decisions a letter home depends on, I write a weekly newsletter for school leaders.
Sources and further reading
- Claire Kelley and Samantha Holquist, "Parents Trust School Communications About Education Far More Than AI, Social Media, or News," Child Trends, July 28, 2026. childtrends.org
- National Parents Union, February 2026 Survey Topline (Echelon Insights, fielded February 12 to 18, 2026). nationalparentsunion.org
- Linea Harding, Michelle Croft, Mark Baxter, Amy Chen Kulesa, Mary K. Wells and Juliet Squire, "Agency in the Algorithm: How K-12 Districts and Advocacy Organizations Can Build Parent Voice on Artificial Intelligence," Bellwether, August 2026. bellwether.org
- Jared Hayden, Michael Toscano and Ken Burchfiel, "AI in the Classroom: Majority of Parents in a New Poll Want a Pause," Institute for Family Studies, September 9, 2026. ifstudies.org
- Jennifer Vilcarino, "Educators Feel Confident They Can Teach About AI. What Do Parents Think?," Education Week, September 9, 2026. edweek.org
- Christopher Joseph Doss, Robert Bozick, Heather L. Schwartz and others, "AI Use in Schools Is Quickly Increasing but Guidance Lags Behind: Findings from the RAND Survey Panels," RAND, September 30, 2025. rand.org
- TeachAI, "Sample Letter to Parents and Guardians," AI Guidance for Schools Toolkit. teachai.org
Dan Fitzpatrick is the founder of The AI Educator, a Forbes contributor and the author of bestselling books on AI in education. More about Dan.
