How to Write a Christian School AI Policy for Coursework

TL;DR

• Most Christian schools still do not have a written policy on how students may use AI in their schoolwork. Teachers are making it up class by class, and students notice the inconsistency before parents do.
• A ban is the easiest policy to write and the hardest to enforce. A better model sorts AI use into three tiers: permitted, permitted with disclosure, and prohibited, with the teacher of record deciding which tier applies to each assignment.
• Ground the policy in what you already believe about honesty and work, extend your existing academic integrity code rather than writing a new one, and apply the same standard to the curriculum vendors you buy from as you do to students.

Ask ten teachers at a small Christian school what the rule is on students using AI for homework, and you will often get ten answers. One teacher has banned it. One requires it for brainstorming. One has not thought about it. The students know exactly which classroom is which, and they adjust.

That is not a technology problem. It is a policy gap, and it belongs to the head of school.

This post is for the principal or head of school who needs to close that gap this year. It follows on from how to present a new elective program to your school board, where I suggested proposing an AI-in-coursework policy alongside any new program. Here is how to write one.

Why does a Christian school need its own AI policy?

Because the questions students are asking are not only technical. They are moral, and your school already has a framework for answering moral questions.

A survey of Christian schools reported by Cardus found that fewer than half of responding schools had a student AI policy in place, while a smaller share had banned student AI use outright. In practice, that means most students in Christian schools are using these tools under rules nobody wrote down.

A written policy does three things that classroom-by-classroom rules cannot:

• It makes the standard consistent, so a student is not judged by one rule in English and another in history.
• It protects your teachers, who otherwise carry the burden of enforcing a rule they invented alone.
• It tells parents what the school believes, which matters more in a Christian school than almost anywhere else. Families chose you partly for how you form character, not just for what you teach.

The good news is that you do not need to start from nothing. Your school already teaches that honesty matters, that work has dignity, and that a student is responsible for what they put their name on. An AI policy is mostly those convictions applied to a new tool.

Why a ban usually fails

A ban feels clean. It is also the policy most likely to be ignored.

AI tools are built into the search engines, phones, and writing software students already use. A ban that cannot be enforced teaches students that rules are suggestions, which is the opposite of what an integrity policy is for. Filters help on school devices, but most homework is not done on school devices.

There is also a formation argument. Students will use these tools in college and at work. A school that only says "no" leaves them to learn the ethics of AI from whoever they meet next. A school that teaches when it is honest to use AI, and when it is not, is doing what Christian education says it does: preparing students to act well in the world they will actually live in.

None of this means AI belongs everywhere. Some assignments exist precisely so students do the thinking themselves. The policy's job is to say which ones.

The three-tier model

The clearest policies we have seen sort AI use into three tiers and let the teacher assign a tier to each piece of work.

Tier 1: Permitted. Students may use AI freely, the way they would use a dictionary or a calculator. Typical uses: checking grammar, generating practice questions, getting a concept explained a second way.

Tier 2: Permitted with disclosure. Students may use AI, but they must say how. A short note at the bottom of the assignment is enough: what tool, what they asked, and what they changed. Typical uses: brainstorming topics, outlining, getting feedback on a draft.

Tier 3: Prohibited. The work must be the student's own, start to finish. Typical uses: in-class essays, tests, personal reflections, and any assignment meant to show what the student can do without help.

Then add three rules that apply across every tier:

1. The student owns the work. Whatever AI touched, the student is responsible for every claim in it and must be able to explain it.
2. No fabrication. Inventing sources, quotes, or data is dishonest whether a person or a tool produced it.
3. When in doubt, ask. If the tier is not stated, the default is Tier 3.

This model works because it puts the decision with the person closest to the learning goal. The teacher of record knows whether a given assignment exists to build a skill or to demonstrate one. For courses where the school relies on a supervising teacher rather than a subject specialist, the same principle holds. See teacher of record for AI-delivered courses for how that role works.

Extend your integrity code instead of writing a new one

The fastest way to get a policy adopted is to make it an amendment, not a new document.

Pull out your current academic honesty or plagiarism policy. It almost certainly already says that students may not submit someone else's work as their own. Add a short section that says AI-generated content counts as "someone else's work" unless the assignment's tier allows it and the student follows the disclosure rule.

Then update the places that policy already lives:

• The student and parent handbook, with the three tiers and two or three examples of each
• The acceptable use agreement students sign for school devices and accounts
• Teacher syllabi, with a line on each assignment or unit naming its tier
• The consequences ladder you already use for plagiarism, so AI violations are handled the same way

Keep the student-facing version to one page. A policy students cannot remember is a policy they cannot follow.

Hold your curriculum to the same standard

Here is the part most AI policies leave out. Your students are not the only ones using AI in your school. Some of the curriculum you buy now has AI built into it, and your policy should say what you expect from those tools too.

If a course includes an AI tutor, ask the vendor plainly:

• Does the tutor give answers, or does it make students explain their reasoning? A tutor that writes the student's response defeats your Tier 3 assignments.
• Can students skip ahead without doing the work? If quizzes and projects can be bypassed, the course is quietly teaching the opposite of your policy.
• Who grades, and can a teacher override the grade? The school, not the software, should own the transcript.
• What student data does the tool collect, and who can see it? Get the answer in writing.

These questions belong in your vendor review. I covered the broader version in how a head of school evaluates an online curriculum vendor and the contract side in what to ask before signing an edtech contract.

For what it is worth, this is how we built Elective Genius. Meri, the tutor inside each lesson, is designed to ask students to explain their thinking rather than hand them answers. Students must pass earlier unit quizzes and submit their projects before the next unit opens. The supervising teacher sees the work and can adjust a score. We run our own school on it, so these rules get tested on real students every day.

A one-page draft you can adapt

Here is a starting point. Change the language to fit your school's statement of faith and handbook voice.

Use of Artificial Intelligence in Coursework

[School name] teaches students to do honest work and to take responsibility for what they put their name on. These convictions apply to artificial intelligence tools as they do to every other resource.

Each assignment will state one of three levels. Permitted: AI tools may be used freely. Permitted with disclosure: AI tools may be used, and the student must note which tool was used, for what, and what the student changed. Prohibited: the work must be entirely the student's own. If no level is stated, the assignment is Prohibited.

At every level, the student is responsible for the accuracy of the work and must be able to explain it. Fabricating sources, quotations, or data is a violation of academic integrity regardless of how it was produced.

Violations are handled under the school's existing academic integrity policy. Questions about whether a use is allowed should go to the teacher before the work is submitted.

That is the whole thing. Teachers will add assignment-level detail. The policy only needs to set the frame.

How to roll it out

A policy that appears in the handbook in August and is never mentioned again will not change anything. A short rollout helps:

1. Draft with two or three teachers, including your most skeptical one. Their objections will make the policy better.
2. Bring it to the board if your bylaws treat academic integrity as policy. Many do. Pair it with any new program that involves AI, so trustees see the guardrails and the program together.
3. Explain it to students in person. Walk through one example of each tier. Ten minutes in chapel or advisory is enough.
4. Send parents a one-page summary with the three tiers and a contact for questions.
5. Review it each spring. The tools will change. Your convictions will not, but the examples should.

Frequently asked

Should a Christian school ban AI for students? A ban is defensible for some assignments and some grades, especially younger students. As a school-wide rule it is hard to enforce and misses the chance to teach students how to use these tools honestly. The three-tier model lets you prohibit AI where it matters most without pretending it does not exist.

How do teachers know whether a student used AI? Detection software is unreliable and should not be the basis of a discipline decision. Better signals come from knowing your students' work, asking them to explain what they wrote, and assigning more Tier 3 work in class.

Does the policy apply to online courses the school provides? Yes. The same tiers and the same integrity standard should apply whether a course is taught in person or delivered online. That is also a reason to review how any online course's tutor and gating work before you adopt it.

Who should own the policy? The head of school, with input from faculty. The board approves it if your governance documents treat it as policy. Teachers apply it assignment by assignment.

About the author

Steve Smith is the founder of Elective Genius and spent 25 years in education before building it. He also runs Rosedale Christian Academy, where every middle- and high-school student uses the platform, which means every decision here gets tested on his own students first. Reach him at schools@electivegenius.com.

Start your pilot

If you are adding electives this year and want them built to the same standard your AI policy sets, start small. One period, one supervising teacher, one semester, and your own students to show you whether it works.

Start Your Pilot or book a school demo.