This guide is for administrators responsible for giving teachers usable AI guidance. It covers how teacher-facing guidelines differ from the board-adopted policy, how to handle the per-assignment decision, what training expectations to set, and how to respond when a teacher suspects AI misuse.

Why teachers need written guidelines

Without written guidelines, teachers apply different rules to the same assignment across classrooms. Parents and students notice the inconsistency. The first time an integrity dispute escalates to the principal's office, the absence of written guidance becomes the center of the conversation.

Written guidelines do three things: they make enforcement consistent, they give teachers cover when they apply the rules, and they surface the district's stance in a form staff can actually reference. The guidelines are a translation of the adopted policy into plain language — not a replacement for it.

Teacher use vs student use

District guidelines have to address both. Teacher use of AI for lesson planning, differentiation, and accommodation is commonly encouraged and represents a different risk profile than student use. Student use is where academic integrity, disclosure, and grade-band decisions apply.

A typical district stance: teachers may use approved tools for lesson prep, rubric drafting, and accessibility scaffolds without disclosure; student-facing work (feedback, grades, communications) requires teacher review before any AI-generated content reaches the student or family. The two uses need separate rules or they conflate.

Per-assignment decisioning

The district sets the gradient; the teacher decides per assignment within that gradient. A useful framing for teachers: "What is the learning I want to protect?" If the goal is the student's own thinking, restrict AI. If the goal is production or accessibility, allow it with disclosure.

Guidelines should give teachers a short decision pattern: identify the target skill, classify the assignment (restricted, allowed with disclosure, or AI-integrated), communicate the classification to students in writing, and hold the line consistently. The classification is explicit because ambiguous classrooms generate disputes.

Training expectations

Districts typically choose one of three training postures: required before first use, required annually, or encouraged but not required. The posture should match the district's overall risk tolerance and capacity. Required-before-use is the safest for adoption but adds onboarding friction. Annual is the most common posture for districts with existing professional development cycles.

Training content should cover: approved tools and their permitted uses, what constitutes PII and what cannot go into a prompt, the district disclosure standard, and the procedure for handling suspected student misuse. Keep it under 90 minutes for the first pass.

Privacy in prompts

The single most important teacher-facing privacy rule is about prompts. Teachers should not paste student names, grades, IEPs, behavior records, or other identifiable information into any AI tool that has not been approved for that data sensitivity level. Aggregating "my class" into anonymized context is usually fine; naming individual students is not.

Guidelines should give teachers concrete substitutions: use "Student A" instead of names, summarize IEP strategies by category rather than quoting, and keep any behavior context anonymous. When in doubt, the prompt should be shareable in a faculty meeting without concern.

Handling suspected misuse

AI-detection tools have high false-positive rates. Teacher guidelines should explicitly warn against using detection output as the sole basis for an academic integrity finding. The recommended procedure is a conference with the student, a look at drafting history (revision records, browser history if available in the LMS), and application of the district's existing academic integrity response.

Suspected misuse is handled under the district's existing discipline framework — not a new AI-specific one. Guidelines should name that framework and cite the relevant sections of the student handbook, so the teacher has one document to reference rather than three.

Onboarding new teachers

New teachers need the guidelines in the first week. Most districts fold them into new-hire onboarding alongside acceptable-use policy, FERPA training, and the student code of conduct. A one-page quick-reference printed and posted in staff workrooms, plus the full guidelines in the staff handbook, covers the dual need for glance-ability and reference.

Adoption checklist

Ten steps to move district AI guidelines from a handbook file into how teachers actually work.

1. Publish the guidelines in the staff handbook. Do not rely on people finding an email from six months ago.
2. Print and post the one-page quick-reference in staff workrooms.
3. Hold a 45-minute faculty meeting introducing the guidelines, with Q&A.
4. Identify one or two teacher champions per building to answer first-line questions.
5. Schedule training — required before first use, annually, or encouraged, per the district’s posture.
6. Pair the guidelines with the approved tools list and a process for requesting new tools.
7. Put one AI case in every first-month staff meeting for the semester. Normalize the decisioning pattern.
8. Give department chairs a subject-specific addendum. Use their language.
9. Document a referral path for suspected student misuse. Route through the existing academic integrity process.
10. Review at semester end. Adjust from real incidents, not hypotheticals.