Set family rules for chatbot homework: help without doing the work

When a child asks, “Can I use a chatbot for my homework?”, saying “yes, but use it responsibly” sounds sensible but does not answer the next question. Can they ask what a word means? Ask for an outline? If the chatbot writes a complete paragraph and the student changes a few words, is that still their work?

A useful family rule needs to be concrete enough that a parent and child can look at the same task and make the same decision. The goal is not to ban AI. It is to keep the part a student is supposed to demonstrate in the student’s hands while allowing help around that work when the teacher and school permit it.

UNICEF reported on June 30, 2026 that analysis across 10 countries estimated at least 20 million children had used AI, including about 13 million who said they used it for learning and homework. Those figures are not Vietnam-specific, but they make the family question practical rather than hypothetical. UNESCO’s guidance on generative AI in education also emphasizes a human-centered approach, privacy and age-appropriate use.

The homework situations below are constructed examples for explanation, not observed outputs from a particular chatbot, account or school.

Start with what the assignment is testing

Imagine this assignment:

Read a 300-word passage about planting trees in cities. Write about 120 words explaining two benefits of urban trees, using at least two details from the passage as evidence.

Before opening a chatbot, underline what the teacher is probably trying to see: Can the student choose two benefits, find supporting evidence, and explain how the evidence supports each idea?

If a chatbot does all three jobs and the student only copies or lightly rewrites the answer, the final paragraph may look polished while hiding whether the student can do the thinking the assignment is meant to assess. If the learning goal is unclear, ask the teacher instead of guessing.

Make three zones on one sheet of paper

Divide a page into three columns.

Zone	When it applies	Example
Okay to ask	AI explains a term or process without producing the work to submit	“Explain ‘urban green space’ in language a seventh grader can understand. Do not write my assignment.”
Do it yourself first	The student needs to produce an initial answer so their own thinking remains visible	Choose the two benefits and evidence first, then ask: “Which sentence in my draft does not clearly connect the evidence to my point? Do not rewrite it for me.”
Stop / do not give it to the chatbot	AI would replace the skill being assessed, or the request contains data that should not be shared	“Write the best 120-word answer for me to submit.”

These zones are not labels for “good” and “bad” AI use. The same kind of help might be allowed during practice and forbidden during a test. Teacher and school rules come first.

Check whether the student can reproduce the thinking

Before submission, a better question than “Did you use AI?” is whether the child can explain the important parts without looking at the chatbot. Ask:

1. What two benefits did you choose?
2. Which sentence in the reading supports each one?
3. Why does that evidence support your point?
4. Is there any sentence in the final answer that you cannot explain yourself?

If the student can answer and the earlier draft shows that the core ideas came from them, AI is more likely to be supporting learning rather than concealing it. If they cannot explain a sentence they plan to submit, stop there, return to the source material, and rewrite it from their own understanding.

This avoids an argument about what percentage of a paragraph was “written by AI.” At the kitchen table, the more useful question is what ability the assignment is assessing and whether the student can still demonstrate that ability.

Add one rule for children’s data

The family sheet should include a rule that applies across subjects:

Do not paste a child’s full name, school or class, address, phone number, passwords, identifiable health information, identifiable grades, or another person’s private material into a chatbot just to finish homework faster.

UNESCO notes that using generative AI can involve sharing data with providers and recommends protecting personal data and considering age-appropriate use. Parents should also check the current age terms of the specific service and the rules that apply where they live. A discussion in UNESCO guidance about age thresholds is not a statement that every service has the same terms or that one age is a universal legal rule.

If an assignment contains real personal information, make a fictional or de-identified version when the teacher permits it, or do not use an external service.

A five-minute family rule sheet

You can start with this and adapt it to the school:

CHATBOT RULES FOR HOMEWORK

1. Before asking AI, say what skill the assignment is testing.
2. OKAY TO ASK: terms, explanations, extra practice questions — when the teacher allows it.
3. DO IT YOURSELF FIRST: ideas, evidence, calculations, or a first draft that the teacher needs to assess.
4. STOP: if the chatbot would produce the answer to submit, the school forbids AI, or the request contains private information.
5. Before submitting, explain the important parts without looking at the chatbot.

The sheet does not need to settle every future AI question. It only needs to make the next homework decision clearer and keep the learning goal visible. Review it with the child after a week: if a rule is too vague to decide a real situation, rewrite that rule rather than adding more abstract warnings.

Sources

• UNICEF Vietnam — Parenting in the age of AI — practical parent-facing guidance on exploring AI with children and discussing benefits and risks.
• UNICEF — Children are adopting AI technologies more than three times faster than adults — June 30, 2026 release supporting the cited cross-country usage figures.
• UNICEF Innocenti — Snapshot of AI usage among children and parents — background for the usage analysis.
• UNESCO — Guidance for generative AI in education and research — supports the human-centered, privacy and age-appropriate-use principles discussed above.