AI Acceptable Use Policy Template for Small Teams

Full AI acceptable use policy

Policy

When to use: The main policy for staff, contractors and volunteers. Share it as a document everyone can find.

{org} AI Acceptable Use Policy

Effective from: {effective_date}
Policy owner: {contact}
Next review: {review_date}

1. Purpose

AI tools can help us write, research, plan and create faster. This policy explains how to use them safely, so that we protect the people we work with, keep our work accurate and stay honest about how it was made. It is meant to help you use AI with confidence, not to stop you using it.

2. Who this policy covers

This policy applies to everyone who works for or with {org}, including employees, contractors, interns and volunteers. It covers any AI tool used for {org} work, on any device, whether the account is paid for by {org} or by you personally.

3. Approved tools

You may use the following AI tools for {org} work:
{approved_tools}

Use your work account where one is provided. Do not use a personal account for work that involves internal information.

If you want to use a tool that is not on this list, ask {approver} first. This includes browser extensions, AI meeting note takers, phone apps and features that connect an AI tool to your email, calendar or shared files. New tools are approved when we understand where the data goes, who can see it and whether it is used to train the provider's models.

4. What information you may share with AI tools

We sort information into three groups.

Public: information that is already published or meant for anyone, such as our website copy, published reports and public job adverts. You may use this with any approved tool.

Internal: everyday working information that is not secret but is not public, such as draft plans, internal process notes and anonymised data. You may use this only with approved tools on a work account, with the privacy settings described in section 8.

Never share: information that could harm a person or {org} if it leaked. Do not paste, upload, dictate or connect any of the following to an AI tool, unless {approver} has approved that specific tool and use in writing:
- passwords, API keys, access codes and security answers
- personal data about customers, clients, students, beneficiaries or donors, such as names linked to phone numbers, home addresses, email addresses or dates of birth
- identity and financial numbers, such as NIN, BVN, passport numbers, card numbers and bank account details
- health, disability, religion and other sensitive personal information
- staff records, including salaries, appraisals, disciplinary matters and medical notes
- client documents covered by a confidentiality agreement, and unreleased financial results
- {sector_data}

If you are not sure which group something belongs to, treat it as Never share and ask {contact}. Removing names is not always enough, because a person can often be identified from other details.

5. Accuracy and human review

AI tools can produce confident answers that are wrong, including invented facts, figures, quotes, sources and links. You are responsible for anything you send, publish or submit, whether or not AI helped you write it.

Before AI assisted work leaves {org}, a person must:
- check every fact, number, date, name and quotation against a reliable source
- open and read every source or link the AI gives you
- read the whole piece for tone, and make sure it sounds like {org}
- get a qualified person to review anything legal, financial, medical or safety related

Never let an AI tool send emails, post content or make payments on its own unless {approver} has approved that automation and a person checks it regularly.

6. Being open about AI use

{disclosure_rule}

Always follow the client's contract if it says something different. Never present AI images, video or audio of a real person as genuine. Do not clone anyone's voice or create their likeness without their clear permission.

7. Ownership and copyright

Work you create with AI for {org} belongs to {org} in the same way as your other work, as far as the law allows. Copyright rules for AI output differ between countries and are still developing, so:
- do not paste someone else's copyrighted work into an AI tool to rewrite it as our own
- check the terms of each tool before using its output in paid client work
- keep a note of which important pieces were made with AI help

8. Accounts, privacy settings and security

- Use a strong, unique password and turn on two step verification for every AI account.
- Where a tool offers a setting that lets the provider use your conversations to train its models, turn that setting off for work accounts.
- Do not share AI accounts or logins between people.
- Do not record a meeting or let an AI note taker join without telling everyone in the meeting first, and stop if anyone objects.
- Delete chats that contain internal information when you no longer need them, where the tool allows it.

9. Fairness and decisions about people

AI must not make final decisions about people. This includes hiring, promotion, pay, grading, discipline, lending, eligibility for services and admission. AI may help organise information, but a person must make and explain the decision. Watch for bias in AI output, for example assumptions about gender, age, ethnicity, religion, accent or disability, and correct it.

10. Reporting problems

Report any of the following to {contact_email} as soon as possible, and within 24 hours:
- you shared Never share information with an AI tool by mistake
- AI produced something harmful, offensive or misleading that reached a client or the public
- you notice an AI tool behaving unexpectedly, such as connecting to accounts it should not
- you are asked to use AI in a way that seems to break this policy

We will treat honest, prompt reports as a chance to fix the problem, not as a reason for blame.

11. Training

Everyone covered by this policy must read it and complete our short AI training {training_deadline}. We will share updates when tools or rules change.

12. Breaking this policy

Breaking this policy may lead to loss of access to AI tools and, for serious or repeated breaches, action under our usual disciplinary process. Hiding a breach is treated more seriously than reporting one.

13. Review

{contact} reviews this policy at least every six months and whenever we adopt a major new tool. The next review is due on {review_date}. Questions go to {contact_email}.

School and classroom AI policy

School

When to use: For schools, tutorial centres and training programmes. Use it with the staff policy for teachers.

{org} AI Use Policy for Students and Teachers

Effective from: {effective_date}
Policy owner: {contact}
Next review: {review_date}

1. Why we have this policy

AI tools can help students understand difficult topics, practise and get feedback. Used badly, they can replace the thinking that learning depends on. This policy explains when AI helps learning at {org} and when it does not.

2. Approved tools

Students and teachers may use these AI tools for school work:
{approved_tools}

Each AI tool sets its own minimum age in its terms of use. Students must only use tools they are old enough to use, and younger students should only use AI with a teacher present.

3. How students may use AI

Unless a teacher says otherwise for a task, students MAY use AI to:
- explain a topic in simpler words, or with a new example
- quiz themselves and practise for tests
- get feedback on a draft they wrote themselves
- check their grammar and spelling
- brainstorm ideas before starting their own work

Students MAY NOT use AI to:
- write all or most of an assignment, essay or project they hand in as their own
- answer questions in a test or exam, unless the teacher allows it
- invent sources, quotes, data or references
- create images, audio or video of another student or teacher

4. Every task has an AI level

Teachers will mark each assignment with one of these levels:
- No AI: the work must be entirely your own.
- AI for planning: you may use AI for ideas and feedback, but every sentence you submit must be yours.
- AI allowed: you may use AI as the task explains, and you must say how you used it.

If a task has no level, ask your teacher before using AI.

5. Saying how you used AI

When you use AI on any assessed work, add a short note at the end, for example: "I used Gemini to explain photosynthesis and to check my spelling. All the writing is my own." Using AI without saying so, where a note is required, is treated as academic dishonesty.

6. Keeping personal information safe

Students and teachers must never put the following into an AI tool: full names with other personal details, home addresses, phone numbers, photos of students, grades, medical or special needs information, or {sector_data}.

7. Rules for teachers

Teachers may use approved AI tools to plan lessons, create practice questions and draft feedback. Teachers must:
- check all AI generated material for accuracy before giving it to students
- never paste student work that contains personal details into an AI tool
- make the final decision on every mark and every report, without relying on AI
- not rely on AI detection tools alone to accuse a student, because these tools can be wrong; talk to the student and look at their drafts first
- tell students and parents when AI was used to create learning materials or feedback, following this rule: {disclosure_rule}

8. Problems and questions

Report any problem, such as personal information shared by mistake or harmful content, to {contact_email}. Questions about this policy go to {contact}. This policy will be reviewed by {review_date}.

One page summary to sign

Summary

When to use: The short version people will actually read. Ask everyone to sign it or reply to confirm.

{org} AI Rules: One Page Summary

1. Use only approved tools: {approved_tools}. Ask {approver} before trying anything new, including extensions and meeting note takers.

2. Never put these into AI: passwords and keys; personal details of customers, students or staff; NIN, BVN, bank and card numbers; health information; confidential client documents; {sector_data}.

3. Check everything. AI can invent facts, figures and sources. You are responsible for what you send or publish.

4. Be open. {disclosure_rule}

5. People decide about people. AI never makes the final call on hiring, pay, grades, discipline or eligibility.

6. Keep accounts safe. Use two step verification, do not share logins and turn off training on your chats where the tool allows it.

7. Report mistakes fast. Email {contact_email} within 24 hours. Honest reports are welcome.

Full policy: {policy_link}
Policy owner: {contact}

I have read and understood the {org} AI Acceptable Use Policy.

Name: ______________________________
Signature: __________________________
Date: _______________________________

Announcement email to staff

Announcement

When to use: Send this when the policy starts, so it lands as help rather than a warning.

Subject: Our new AI rules, and a short briefing on {session_time}

Hi everyone,

Many of us now use AI tools to write, research and plan, and that is a good thing. To make sure we use them safely, we have put together a short AI policy for {org}. It starts on {effective_date}.

The short version:
- We have approved these tools: {approved_tools}.
- Some information must never go into an AI tool, such as customer personal details, passwords and confidential client documents.
- A person checks everything AI helps us produce before it goes out.
- If something goes wrong, tell {contact_email} quickly. Honest mistakes reported early are easy to fix.

You can read the full policy here: {policy_link}

We will go through it together on {session_time}. It will take about 30 minutes, and you are welcome to bring questions or tools you would like us to approve. Please read the one page summary and confirm you have read it {training_deadline}.

Thank you for helping us use AI well.

{sender_name}