Olsen & Sons
← Insights

AI · 4 min read

A one-page AI policy for a small business

Your team is already using AI. Here is how to make sure they use it safely.

By David Olsen · November 21, 2026

If you haven't told your team how to use AI, they are deciding for themselves. Some are pasting customer information into free tools. Others are afraid to touch it. Neither is good for the business.

You don't need a twenty-page document. You need one page that answers five questions.

1. Which tools are approved?

Name them. Pay for business or team plans; they come with better privacy terms and keep your information out of model training. Free personal accounts are where most leaks happen.

2. What may never go in?

Be specific: customer financial information, health information, passwords, employee records, anything a client gave you in confidence, and anything covered by a contract. When in doubt, leave it out or remove the identifying details.

3. Who checks the output?

AI can be confidently wrong. Anything that goes to a customer, a regulator or a court is read and approved by a person, who owns it.

4. When do we tell people?

Decide when customers should know AI was involved, for example in reports or recommendations, and be consistent.

5. Who do I ask?

Name one person who answers questions and keeps the approved list current.

Then train

A policy nobody understands doesn't protect anyone. A half-day of hands-on training on your own work does more than any document. That is the training we give.

See what we'd change first.

Book a free Walkthrough. We spend about an hour seeing how your team works, and within two business days you get a one-page memo with the three changes we'd make first. It's yours whether or not you hire us.