Guide

AI use policy template for small businesses (free, 2026)

What a workable AI use policy contains, a short example you can copy, and how to roll it out to your team in one week.

Updated 2026-11-19

Your people already use ChatGPT, Copilot, Gemini or DeepL at work, whether or not anyone wrote rules for it. An AI use policy turns that into something you control: which tools are fine, what never goes into them, and who to ask. For a small business it should fit on two pages.

What a good AI use policy contains

  1. Scope. Who it applies to: employees, freelancers and anyone working on your behalf.
  2. Approved tools. A short list of the AI tools people may use for work, ideally business accounts with training on your data switched off.
  3. What never goes in. Passwords, customer personal data, health data, bank details, unreleased financials, anything confidential, unless a tool is explicitly approved for it.
  4. Check before you use. AI output can be wrong or invented. Facts, figures, names and quotes are checked by a person.
  5. Responsibility. The person who sends, publishes or decides is responsible, whether or not AI helped.
  6. People decisions. AI is never the only basis for a decision about a customer, applicant or colleague.
  7. Transparency. Tell people when they talk to a chatbot, and before a meeting is recorded or transcribed.
  8. A contact person. One named person approves new tools and answers questions.

A short example

Rule 1. Use only the AI tools on our approved list, with the company account.
Rule 2. Never paste customer data, passwords or confidential documents into an AI tool unless it is approved for that data.
Rule 3. Treat AI output as a draft. Check every fact before it leaves the company.
Rule 4. Ask our AI contact person before using a new tool.

How to roll it out in one week

  • Day 1: list the AI tools people actually use. Ask; the answer is usually longer than you think.
  • Day 2: decide which ones are approved and on which account.
  • Day 3: write the policy with your company name, tools and contact person.
  • Day 4–5: share it and run a short training so people understand the why, not just the rules.
  • Then: keep a simple record of who was trained and review the policy once a year.

If you are in the EU

Since February 2025, Article 4 of the EU AI Act asks companies that use AI tools to take measures for the AI literacy of their staff. A written policy plus a short, documented training is a practical way to show you did. Read our guide on what Article 4 asks of small businesses.

Common mistakes

  • Banning AI outright. People use it anyway, on private accounts, where you have no control.
  • A 20-page policy nobody reads. Short rules with examples work better.
  • No owner. Without a named contact, new tools appear without anyone checking them.

You can write the policy by hand, or let our free generator fill in your company name and tools in five minutes.

Make your AI policy in 5 minutes

The free generator writes a policy with your company name and tools. The full kit adds the tool register, the training plan and a 15-minute staff training with certificates.

More guides