Guide

What a school AI policy should include

A useful policy connects AI use to learning goals, gives teachers clear responsibilities, and establishes expectations students can understand.

A school AI policy needs to answer a question that broad statements about responsible use often leave unresolved: what must students be able to do independently, and when may they use assistance?

Without that distinction, teachers are left to interpret general rules assignment by assignment. Students encounter inconsistent expectations, and disagreements about AI use become difficult to resolve. The policy should provide enough direction to support sound decisions while leaving room for differences across subjects, grade levels, and learning needs.

Connect expectations to the assignment

The starting point is the learning objective. What is the assignment intended to teach or assess? What evidence will show that the student has learned it? Which uses of AI would support that learning, and which would allow the student to bypass it?

For a research assignment focused on source evaluation, students may need to locate, assess, and justify their sources independently. In another assignment, they might critique an AI-generated summary against the original text. Both involve research, but they require different rules because they assess different skills.

Schoolwide expectations should establish a consistent framework. Teachers then apply it to the work students are actually doing.

Include the provisions people will need

  • Scope. Identify who the policy covers, including students, teachers, staff, and contractors. Address instructional and administrative uses, including assessments, IEP documentation, family communication, and letters of recommendation.
  • Permitted use. Explain how teachers will specify acceptable assistance with research, drafting, revision, practice, and assessment. Give examples that help students distinguish permitted support from work they must complete themselves.
  • Disclosure. State when students and staff must acknowledge AI use and what that acknowledgment should include. A brief description may be sufficient for some tasks; others may require relevant prompts and outputs.
  • Assessment. Identify when independent performance is required and how it will be demonstrated. Account for approved accommodations and assistive technologies.
  • Staff responsibilities. Set expectations for planning, instructional materials, feedback, grading, and communication. Specify where professional review is required and who remains responsible for accuracy and final decisions.
  • Student information and approved tools. Explain what information may be entered into which systems, what must be excluded, and how new tools are reviewed and approved.
  • Access. Ensure that assignments do not depend on students purchasing AI tools or having access that the school cannot provide.
  • Review. Name the person or group responsible for reviewing the policy, set a review date, and provide a way for staff, students, and families to raise concerns.

Avoid approaches that undermine trust

In his focus groups with high school students, Tony Frontier identified concerns about blanket bans, inconsistent expectations, and inaccurate claims about AI's capabilities. These concerns deserve attention because they affect whether students regard a policy as credible and fair.

A blanket ban can be difficult to enforce and may discourage students from asking legitimate questions about acceptable use. Restrictions are easier to explain when they are tied to a specific learning purpose, such as demonstrating independent writing or mathematical reasoning.

Silence creates its own problems. When a school provides no common guidance, students must navigate differences that may have little educational justification. Teachers need discretion, but students should be able to understand why expectations differ.

Staff also need to be accurate about what AI can and cannot do. Overstated claims weaken otherwise reasonable guidance. Teachers do not need to be experts in every tool; they do need to distinguish established limitations from assumptions and acknowledge uncertainty when it matters.

Assign decisions to the appropriate level

The board establishes the governing principles and schoolwide requirements, including expectations for privacy, academic integrity, and fair treatment. Administration translates those requirements into procedures, tool approvals, staff training, and communication with families.

Teachers determine acceptable use for individual assignments within that framework. They should communicate those expectations before work begins, with examples where the distinction may be unclear.

Students should not have to infer a rule from a general statement about honesty or discover it after submitting their work.

Establish a fair process for suspected misuse

An AI detector score should not, by itself, determine a disciplinary outcome. Detection tools can produce false positives, and their results do not establish how a student completed an assignment.

When concerns arise, teachers should consider evidence gathered through the normal learning process: drafts, source notes, revision history where available, and conversations about the work. A student's explanation can help clarify their understanding, but difficulty answering a question is not proof of misconduct.

The policy should explain how concerns are raised, how students can respond, and how decisions are reviewed. Consequences should reflect the circumstances, including whether expectations were clear and whether the student had been taught how to follow them.

Test the policy in practice

Begin with a course or unit and work through its assignments. Write the AI expectations, discuss them with students, and examine where questions remain. Can students distinguish acceptable help from prohibited assistance? Can teachers apply the rules consistently? Does the assessment still provide useful evidence of learning?

Use those findings to revise the policy before extending it across the school. Schedule a further review after implementation, when teachers and students can identify provisions that are unclear or difficult to apply.

A policy earns its value through these everyday decisions: students understanding what is expected, teachers being able to explain why, and school leaders having a fair basis for resolving concerns.

Working on one?

We develop AI policy for schools and districts, including the supporting documents and a board presentation when the engagement includes one.