AI Policy for Schools: A Template for Clear Expectations

An AI policy for schools sets the conditions for using AI in teaching, schoolwork, and staff duties. It should name the learning purpose, allowed uses, disclosure requirements, privacy boundaries, and person responsible for review. Students need an answer they can use before starting an assignment, while teachers need room to protect the skill being assessed.

Tools like fenbs can keep approved staff rules and the reasons behind them available to connected assistants. Start with the policy itself: an assistant having access to instructions does not make a classroom tool approved or make its output correct.

What should the policy help your school decide?

Begin with learning, rather than a list of products. Name the skills students must demonstrate independently and the activities where supervised assistance could help. A spelling exercise and a history project may need different boundaries, even when students use the same device.

Separate the schoolwide policy from assignment instructions. Schoolwide rules cover approved tools, privacy, honest attribution, and review. Each assignment then states whether students may use AI for that task, for which steps, and what evidence of their own thinking they must submit.

For the broader structure outside education, see the general AI policy guide. Keep a named school policy owner, an effective date, and a review date at the top of your version so families can identify the current expectations.

For drafting methods, see AI for policy writing and the AI policy generator guide. AI in policy making addresses public-policy work, while AI policies for nonprofits addresses nonprofit operations.

Which clauses can you adapt for a first draft?

The following is sample language to adapt, not an adopted policy. Have your district privacy lead, instructional leaders, and counsel align it with state requirements, vendor agreements, student ages, and existing academic integrity procedures. Replace every general reference to approval with the actual person or process students should use.

• Purpose: We use AI only when it supports a stated learning or schoolwork goal. Students remain responsible for understanding and explaining the work they submit.
• Permission: Teachers state the permitted AI activities for each assignment. When instructions are unclear, students ask before using AI on assessed work. Permission for one assignment does not carry over to another.
• Approved access: Staff and students use only tools and account arrangements approved for their role and age. A teacher demonstrating a tool does not authorize students to create personal accounts.
• Disclosure: Students identify the tool, the task it helped with, the material they kept, and how they checked it. Teachers explain whether prompts, drafts, citations, or a short use statement must accompany the work.
• Privacy: Do not enter student records, identifying details, private communications, or other restricted information into an AI tool without the school review and authorization required for that use.
• Review: A responsible educator checks AI-assisted instructional materials before use. A person remains responsible for grading decisions, feedback, and decisions affecting student opportunities.
• Access and concerns: Students can request an approved alternative and report an unsafe response or accidental disclosure to the named school contact. Concerns follow existing school procedures, with a chance for students to explain their work.

Keep the family version readable on a phone and available through the channels families already use. Put implementation tasks and unresolved questions beside the approved policy in fenbs Decisions and rules, so a suggestion cannot quietly become a school instruction.

How do teachers make permission clear for each assignment?

Give students a short instruction next to the assignment, including examples of allowed and disallowed help. Avoid saying only that AI is allowed: a student may reasonably interpret that as permission to produce the entire response.

For disclosure wording, distinguish acknowledging help from citing evidence. The MLA guidance on AI citation↗ (opens in a new tab) recommends consulting the sources an AI response links to and citing those sources directly. Your assignment should also explain how students acknowledge planning, editing, or other permitted assistance.

What privacy checks belong before classroom use?

FERPA, the federal student education records privacy law, applies to educational agencies and institutions receiving funds under programs administered by the US Department of Education. It generally requires consent before disclosure of personally identifiable information from education records, unless an exception applies. The FERPA regulations↗ (opens in a new tab) set out the conditions; classroom usefulness alone is not an exception.

COPPA concerns certain online services collecting personal information from children under 13. The FTC guidance↗ (opens in a new tab) explains its scope and notice and consent duties. Ask the district to assess the specific tool and use; do not assume a student account, a teacher account, or a family acknowledgment settles every privacy question.

• Identify the information sent, including uploaded files, names, images, voices, and account identifiers.
• Check the vendor terms, permitted ages, access controls, retention, deletion, and any use of data for model training.
• Document the approved educational purpose, authorized users, and any required notices, permissions, or contract terms.
• Keep identifiable student examples and incident details in school-approved systems; use fictional examples in general AI instructions.

How should staff respond to mistakes and suspected misuse?

Specify who receives reports and what happens next. An inaccurate explanation, an inappropriate response, an accidental data disclosure, and suspected dishonest submission need different responses. Stop the affected activity when necessary, preserve evidence in the authorized school system, and route the concern to the appropriate staff member.

For academic concerns, review the assignment instructions and the student explanation alongside available drafts or other work evidence. Use the school academic integrity and appeal procedures. A fair policy should tell students how they can correct a missing disclosure or challenge a concern, rather than leaving that process to an unexpected conversation.

Review the policy after a pilot and at scheduled intervals. Ask whether students understood the instructions, teachers could apply them, and alternatives were practical. Record the reason for an approved change; retain sensitive case details elsewhere. Our guide to using AI at work also helps staff define small, reviewable uses for administrative tasks.

How to do this with fenbs

1. Create a free account to try a policy project with non-sensitive sample material. Use a team kanban board with suitable member roles when approved staff need to work together.
2. Add the policy project and tasks for instructional review, privacy review, family guidance, and teacher examples. Keep the adopted handbook in its official school location.
3. Record human-approved expectations in Decisions and rules. Include their scope and reason; leave unresolved choices open for the responsible school leader.
4. Connect an approved assistant using the relevant Integrations guide, then ask it to read the project context before drafting staff guidance. Check its draft against the adopted policy.
5. Save Where we left off with the next review task and decisions still waiting on a person. Record approved revisions and useful lessons when the pilot ends.

fenbs does not approve education vendors, enforce classroom restrictions, or replace student information systems. It fits the staff work of keeping approved instructions consistent; it is not the right place for sensitive student cases without the required school assessment and authorization.

An AI brain keeps the work record outside a single chat. Explore the AI governance guides when you are ready to connect that record to the wider staff review process.