Your school's AI policy, explained — and the four questions to ask
Most policies ban "unauthorised" AI use without defining authorised use, which pushes the judgement call onto each teacher. Get the answer in writing per class: is AI allowed for brainstorming, explaining, checking, or none of those, and how are accusations handled. Keep drafts and version history — that evidence settles disputes that a detector score cannot.
What these policies usually say
Open your school's handbook and search for "artificial intelligence". Nine times in ten you'll find a single line inside the academic-integrity section: submitting work generated by AI as your own is dishonesty, and unauthorised use of AI tools is prohibited.
That sentence does real work and leaves the important part out. It never defines what authorised use is. Which means the actual rule in your child's life is whatever each individual teacher decided in August.
Suppose the handbook prohibits 'unauthorised assistance,' but tonight's science sheet says students may use reference materials. Don't assume reference materials automatically include an AI explanation. Put the two instructions together and ask the teacher which forms of help are permitted for this task. The useful question names an activity, such as explaining a term or checking calculations, rather than asking whether all AI is allowed.
Imagine an English teacher allows spelling checks but not generated revisions. Your child needs to know what to do when a tool offers a replacement sentence rather than a spelling correction. Ask for that distinction before accepting the suggestion. Keep the answer beside the assignment. A familiar editing interface doesn't make every feature within it part of the permission.
The three policy types you'll meet
- Ban. No AI on any assignment. Rare at district level, common in individual English and history classes.
- Disclose. AI is allowed if the student says how they used it, usually in a note at the end of the assignment. This is now the most common workable version.
- Task-dependent. Allowed for brainstorming, outlining and revision; not allowed for drafting or final text. Best in spirit, hardest to enforce, and the one that produces the most misunderstandings.
The four questions to ask each teacher
- For this class, is AI allowed for brainstorming or outlining?
- Is it allowed for explaining a concept my child didn't understand, or for checking finished work?
- If it is allowed in any form, how should my child disclose it?
- If work is suspected of being AI-generated, what happens — and is a detector score used as evidence?
Send those four in one short email and keep the reply. It takes a teacher two minutes and it converts an ambiguous policy into a rule your child can actually follow.
Suppose you send a short message with the assignment name and a concrete example: may my child ask for an explanation of evaporation using different examples, without submitting generated text? The teacher can answer the particular request. If permission is conditional, record the condition too, such as disclosure or use only during practice. A broad 'yes' shouldn't lose the details that made it a yes.
For disclosure, imagine the permitted activity was practice questions. Ask where the student should note that and what detail is required. The note should describe what actually happened. Calling an outline 'spelling help' isn't accurate, and writing a disclosure doesn't convert prohibited assistance into permitted assistance. Let your child practice wording an honest note before the deadline makes everything feel urgent.
Why the last question matters most
AI detectors produce false positives, and they fall hardest on multilingual students, neurodivergent writers and anyone with a plain, orderly style. A school that treats a percentage score as proof will eventually accuse an honest student. Read are AI detectors accurate before you need it, and flagged for AI: what to do if you already do.
The evidence that actually protects your child
- Draft history. Google Docs and Word both keep version history — it shows work accumulating over time, which no detector can argue with.
- Handwritten notes or an outline photographed before the draft.
- The ability to explain the work out loud. A two-minute conversation beats every automated check.
That last one is the point of the AI homework checker: it doesn't score the work or guess at AI, it tells you what the work suggests your child understood and gives you the questions to ask at the table. If the answer is "not much", explain my kid's homework walks the concept back through at their level.
Suppose your child writes from a paper outline and revises in a document. Keep those genuine materials together with the class instructions and permission email. They can help explain the process if questions arise. They aren't a guarantee about the school's decision, and they shouldn't be altered to make the record look stronger. A simple folder is enough; the goal is an accurate account, not an elaborate defense for every sentence.
Imagine the family reviews the policy after a new assignment introduces group work. Ask which parts each child must do and whether any tool use needs to be disclosed by the group. Keep the home rule consistent with the teacher's answer. When the instructions remain unclear, begin the independent reading or notes and wait for clarification before using the disputed assistance.
Set the household rule above the school rule
School policies vary by class and change every year. A household rule doesn't. Ours is one line: AI may explain, quiz and check — it may not write. Put your version in writing with the family AI agreement template, and use the grade-by-grade rulebook when you need to settle whether a specific kind of help crosses the line.
Frequently asked questions
What is a school AI policy?
It is usually a clause inside the academic integrity policy stating that submitting AI-generated work as your own is dishonesty and that unauthorised AI use is prohibited. Most policies do not define what authorised use is, leaving the decision to individual teachers.
Is using AI for homework against school rules?
It depends on the class. Many schools allow AI for brainstorming, explaining a concept or checking work, but not for producing text that is handed in. Because the handbook rarely spells this out, ask each teacher in writing what is allowed in their class.
Can a school punish my child based on an AI detector score?
Some do, though detector scores are probabilistic and produce false positives, particularly for multilingual and neurodivergent writers. Ask whether a score alone can trigger a penalty, and keep draft version history as counter-evidence.
What should I email my child's teacher about AI?
Ask four things: whether AI is allowed for brainstorming, whether it is allowed for explaining or checking, how use should be disclosed, and what process applies if work is suspected of being AI-generated. Keep the reply.
How do I prove my child wrote their own work?
Document version history in Google Docs or Word, keep handwritten notes and outlines, and be able to have your child explain the reasoning out loud. That combination is far more persuasive than disputing a detector percentage.
Have a specific piece of homework in mind?
Paste it — or a screenshot of it — into the Homework Check on our homepage and ask your question. You'll get a plain-language read on whether the thinking looks like your kid's, what stood out, and the questions worth asking them. First check is free.