School leadership guide
AI policy guidance for schools
A useful school AI policy does more than list tools. It gives students and employees clear boundaries, names who decides, and connects everyday practice to privacy, learning, safety, and accountability.
Written and reviewed September 4, 2026Quick answer
What should a school AI policy cover?
- Purpose and human-centered principles
- Who and what the policy applies to
- Allowed, supervised, and prohibited uses
- Student and employee data protections
- Teaching, assessment, authorship, and disclosure
- Vendor review and procurement
- Human review for consequential decisions
- Reporting, response, training, and scheduled review
Govern uses and risk—not brand names
Product lists become outdated quickly because AI features appear inside tools schools already use. Define expectations for capabilities and situations: generative writing, image or audio creation, automated feedback, predictive recommendations, student-facing chat, transcription, and use of confidential information.
Pair stable principles with shorter operational guidance that can be updated more often. A single policy should not try to replace privacy, security, academic integrity, accessibility, procurement, records, employment, or safeguarding procedures.
Make permission understandable
A three-level model is easier to apply than vague language: allowed uses can proceed under ordinary expectations; supervised uses need teacher, leader, privacy, or technology review; prohibited uses are inconsistent with safety, learning, law, contract, or institutional values.
- Allowed: low-risk brainstorming with no protected data, when the assignment permits it.
- Supervised: student-facing tools, uploading institutional content, automated feedback, or new vendor trials.
- Not allowed: entering highly sensitive records into unapproved systems or delegating final consequential decisions to AI.
Name the decision owners
Specify who can approve classroom experiments, who reviews privacy and security, who owns academic expectations, who can authorize procurement, and what must be escalated. Human oversight is meaningful only when the reviewer understands the context, can challenge the output, has authority to change the result, and accepts responsibility for the final action.
Plan implementation before approval
Not every local policy requires board or superintendent approval. Match the approval route to your governance structure, and record the person or group authorized to approve it.
- Inventory current and embedded AI uses.
- Publish role-specific examples for students, faculty, staff, and leaders.
- Train people before enforcing unfamiliar expectations.
- Provide a question and incident-reporting path.
- Document vendor reviews, decisions, exceptions, and expiration dates.
- Set a review date and revise when evidence, law, contracts, or school needs change.
Recommended learning
Go deeper with a complete course.
The guide gives you a practical starting point. These self-paced courses build the underlying judgment, application, and evidence of learning.
Ethical AI Governance in Schools
A senior leadership and board certificate for governing AI across mission, privacy, learning, procurement, high-stakes decisions, workforce practice, and institutional risk.
View course and enroll →Ethical AI for School Leadership and Operations
A cross-functional certificate course for school leaders and office teams on using AI to increase capacity while preserving privacy, accuracy, institutional trust, relationships, mission, and human accountability.
View course and enroll →Ethical AI Governance in Higher Education
A governance certificate for presidents, provosts, deans, trustees, cabinet members, technology leaders, research leaders, and other higher-education decision-makers.
View course and enroll →Primary sources
Continue with the official guidance.
This guide is educational information and a planning aid. Applicable law, contracts, and local policy should be reviewed by the people responsible for them.
Independent-school educator and technology leader with experience in teaching, faculty development, school operations, responsible AI, and applied educational research.
Guidance is checked against primary sources and revised as technology, research, and institutional expectations change.