Write a short AI use policy for your team: approved tools, the information that never goes into them, how output is checked and who to ask. Print it for each person to sign, without signing up.
RULES BEFORE TOOLSNo account needed
Edit the sample policy for your business, then print it for your team to read and sign. Your draft stays in this tab if you refresh; save a file to keep it. Nothing is uploaded, and this page does not use AI.
1. The business
2. Approved AI tools
Name each tool your team may use for work and what it is for. Leave the list empty if none is approved yet; the policy then tells staff to ask first.
Tool 1
Tool 2
3. Never enter this information
This row is already first.
This row is already last.
4. Using what AI produces
This row is already first.
This row is already last.
5. New tools and sign-off
Coverage check
The never-enter list covers customer details, payment data, staff records and passwords.
Rules on data, hiring and advertising vary by state and industry. This template is not legal advice; see the guidance linked below the tool.
Policy preview
AI use policy
Meadow Corner
Effective
November 2, 2026
Review by
May 3, 2027
Questions to
Alex Morgan, owner
Why we have this policy
AI assistants can save time on drafting and summarizing. This policy sets out which tools we use, what never goes into them and how we check what they produce, so customers, suppliers and the team can trust our work.
Approved AI tools
Business-plan AI chat assistant: Drafting emails, checklists, job ads and product descriptions; summarizing sales exports with customer columns removed.
Spelling and grammar checker: Checking signs, social posts and the website before they go live.
Never enter this information into an AI tool
Customer names, email addresses, phone numbers, addresses or order histories.
Payment card numbers, bank details or anything from the card terminal.
Staff records, pay, availability notes, health information or performance notes.
Passwords, alarm codes, safe combinations or login links.
Supplier prices or terms that we agreed to keep confidential.
Using what AI produces
Read every draft before using it, and check prices, dates, quantities and product facts against our own records.
A person approves anything sent to a customer or supplier, or posted publicly.
Do not use AI to make or rank hiring, scheduling, pay or discipline decisions about people.
Do not publish product claims, such as health or environmental benefits, that we cannot back up.
Report a mistake or anything entered by accident to the policy contact the same day.
Asking for a new tool
Ask the policy contact before trying a new AI tool or turning on an AI feature in software we already use. Say what it is for and what information it would see.
Questions
Ask Alex Morgan, owner about anything this policy does not cover.
I have read this policy and will follow it.
NameSignatureDate
Reopen a JSON file downloaded from this tool, up to 1 MB. Read on your device without uploading.
Before your team pastes anything into an AI assistant, write down the rules. This free AI policy template starts with a sample policy for a small shop: the approved AI tools and what each is for, the information staff must never enter, how output is checked before anyone relies on it, and who answers questions. Edit it for your business, then print a copy for each person to read and sign.
Make an AI policy for your team
Enter your business name and the person staff ask about AI use, usually the owner or manager.
List the approved tools and what each may be used for, such as drafting supplier emails or checklists. Leave the list empty if nothing is approved yet; the policy then tells staff to ask first.
Edit the information staff must never enter. The sample covers customer details, payment data, staff records, passwords and confidential supplier terms.
Edit the rules for using output, such as checking prices and dates against your own records and having a person approve anything sent to a customer.
Add how to ask for a new tool, choose whether to print a signature line, and set an effective date and a review date.
The coverage check notes when the never-enter list leaves out customer details, payment data, staff records or passwords. It is a reminder, and it never stops you printing.
What a workplace AI policy should cover
A short policy people read beats a long one nobody opens. Most small businesses need five parts:
Approved tools. Name each assistant or AI feature and what it is for. Include AI features inside software you already use, such as an email or point-of-sale system.
Data that never goes in. Customer contact details, card and bank data, staff records and passwords are the usual starting list. Read each tool's terms for how it stores what you enter and whether it uses it for training.
Checking output. AI tools can state wrong prices, dates or facts with confidence. A person checks a draft before it reaches a customer, a supplier or the public.
Decisions about people. Keep hiring, scheduling, pay and discipline decisions with a person. In the United States, the EEOC explains which employment practices are prohibited, and the same rules apply when software helps make the decision.
Who to ask and what to report. Name one contact, and ask staff to report a mistake or anything entered by accident the same day.
The NIST AI Risk Management Framework is a voluntary guide to the risks to plan for if you want more depth. Privacy and advertising rules vary by state and industry; California's attorney general explains which businesses the California Consumer Privacy Act covers. This template is not legal advice.
Print, sign and review
Print one copy for each person, or save it as a PDF from the print dialog. Keep the signed copies with your other staff records, and add reading the policy to the onboarding checklist for new starters. Review the policy on the date you set, or sooner when you approve a new tool. Save a draft file to open and edit it here next time.