Interview: A Middle School Teacher Explains Her ‘AI Allowed’ Homework Policy

The policy in one sentence: “AI is allowed, but thinking is not optional”

Parents are getting mixed messages about AI in school. Some teachers ban ChatGPT completely. Others quietly tolerate it. And many are trying to build reasonable middle school AI rules while the technology changes every month.

To get a grounded perspective, I spoke with Ms. Elena Park, a 7th-grade ELA and social studies teacher at a public middle school in the Midwest. This interview breaks down her ai allowed homework policy, what she’s actually seeing with student work, and the ai in classroom guidelines she shares with families.

Her framing is simple: “I’m not grading how well a chatbot writes. I’m grading how well a student thinks, revises, and learns.”

Below, you’ll find her exact rules, a parent-friendly disclosure template, and a practical table you can use tonight.

Why she moved from “No AI” to an AI allowed homework policy

Q: You used to prohibit AI tools. What changed?

Ms. Park: At first I tried a total ban. It felt like the safest option. But the truth is, I couldn’t enforce it consistently—and students knew that. A ban also created a weird dynamic where honest kids were anxious and sneaky use was rewarded.

What pushed me to an “AI allowed” approach was realizing AI can support learning when it’s used like:

• a tutor that helps a student understand directions
• a brainstorming partner for ideas
• a tool for checking clarity and grammar

But it becomes a problem when it replaces the student’s thinking.

Q: How do you explain this to parents who worry it’s “cheating”?

Ms. Park: I compare it to calculators. If a student uses a calculator before they understand the concept, it hurts them. If they use it to check their work after learning, it supports them.

We need to teach kids the difference between:

• outsourcing thinking (not okay)
• supporting thinking (often okay)

Q: What’s your biggest concern right now?

Ms. Park: Kids believing the first answer they get. AI can sound confident and still be wrong. So my main goal is building habits like verifying sources, explaining reasoning, and showing their process.

This is also a big part of how teachers feel about ChatGPT: it’s not just about plagiarism—it’s about students skipping the mental “workout” that grows skills.

Ms. Park’s middle school AI rules (and how she enforces them)

Ms. Park’s policy fits on one page and is repeated on every major assignment. She calls it “Allowed Tools, Required Proof.”

Q: What are your core rules?

Ms. Park: Here are my middle school AI rules, written in student-friendly language:

• You may use AI for: brainstorming, outlining, generating practice questions, defining vocabulary in simpler language, and grammar suggestions.
• You may not use AI for: writing your final paragraphs, answering comprehension questions, or creating citations/sources you did not read.
• You must show your thinking: notes, drafts, quick writes, or a short reflection explaining choices.
• You must disclose AI use: what tool you used and how you used it.

Q: What does “disclose” look like for a 12-year-old?

Ms. Park: I don’t want a scary legal statement. I want a simple, honest note.

Here’s the template I give them:

• Tool used: (ChatGPT, Gemini, Grammarly, etc.)
• What I asked it: (copy/paste one prompt)
• What I used: (ideas, an outline, feedback on clarity)
• What I changed: (2–3 bullets)

Q: How do you catch misuse without turning into a detective?

Ms. Park: I don’t rely on AI detectors—they’re unreliable, especially for multilingual students.

Instead, I design assignments that make the process visible:

• in-class “seed” writing (5–8 minutes) that becomes part of the homework
• checkpoints (outline due before draft)
• a short conference: “Explain why you chose this example”

If a student can’t explain their own work, that’s a learning issue we address.

The practical guidelines: what’s allowed for different homework types

Parents often ask, “Okay, but what does ‘AI allowed’ mean for my kid’s assignment?” Ms. Park shared a simple decision guide she uses.

Homework type	AI allowed?	Allowed uses	Not allowed	What parents can check in 2 minutes
Reading comprehension questions	Limited	Define tough words, rephrase directions	Writing the answers	Ask: “Show me the page where you found that answer.”
Essay or paragraph writing	Yes, with proof	Brainstorm, outline, feedback on clarity	Final drafting, adding quotes they didn’t read	Look for: outline + 1–2 drafts + a reflection note
Research project	Yes, carefully	Topic narrowing, search terms, summarizing notes from real sources	Fake citations, “sources” that don’t exist	Ask: “Open the sources—can you find the quote?”
Math word problems	Limited	Explain steps, generate extra practice problems	Giving final answers without steps	Check: written steps + one self-correction
Studying for a quiz	Yes	Flashcards, practice questions, mini-explanations	Memorizing incorrect info without verification	Have them verify 2 answers in notes/textbook

A key theme in her ai in classroom guidelines is that AI should increase practice, not reduce it.

What parents should say at home (without turning it into a fight)

Q: If you could give parents one script to use, what would it be?

Ms. Park: I’d say:

• “AI can help you learn. It can’t do the learning for you.”
• “If you use a tool, you need to be able to explain the result.”
• “If you can’t describe it in your own words, it’s not ready to turn in.”

Q: What are red flags that a student used AI in a harmful way?

Ms. Park: Things parents notice first are:

• vocabulary that doesn’t match how the child normally speaks
• perfect structure but shallow ideas
• no messy drafts, no notes, no “in between” work
• quotes or facts they can’t locate again

Q: What are green flags that AI use is actually helping?

Ms. Park: I love when students use AI to:

• ask for examples, then write their own
• get feedback and revise (real revision, not just rephrasing)
• generate a practice quiz and check it against class notes

That’s what “AI allowed” is supposed to look like.

Parent tip: turn AI into a teaching moment

If your child uses ChatGPT, try a quick “trust but verify” routine:

• Ask them to read the AI output out loud and underline anything they don’t understand.
• Have them replace one sentence with their own words.
• Ask: “Where did that fact come from?” and make them find the source.

This approach supports learning and aligns with how teachers feel about ChatGPT: helpful when it builds skills, harmful when it replaces them.

Next Steps: set your family’s AI homework rules in 15 minutes

You don’t need a perfect policy. You need a clear one that matches school expectations and helps your child build honest habits.

Here’s a simple plan:

• Step 1: Ask the teacher one question.

  • “What is your AI allowed homework policy, and what do you want students to disclose?”

• Step 2: Pick your home rule (start small).

  • Example: “AI is allowed for brainstorming and explaining directions, but not for writing final answers.”

• Step 3: Use the one-paragraph disclosure note.

  • Have your child paste the prompts they used and list 2 changes they made.

• Step 4: Add one ‘proof of thinking’ requirement.

  • Choose one: an outline, a photo of notes, or a 3-sentence reflection.

• Step 5: Do a 2-minute spot check once a week.

  • Can they explain the work? Can they locate a source? Do they understand the vocabulary they used?

If you set expectations now, your child learns something bigger than any single assignment: how to use powerful tools with integrity. That’s the real goal behind today’s middle school AI rules—and it’s a skill they’ll carry into high school, college, and future jobs.