ChatGPT at work: a simple AI usage policy for your team

If your team writes emails, proposals or social posts, some of them are almost certainly using ChatGPT, Gemini or similar tools already. That's not a problem in itself. The risk is what they paste in, and what they send out without checking.

A short, clear policy fixes most of that. Here's a template you can adapt.

Why not just ban it?

Bans rarely work. People use the tools on their own phones instead, and you lose any visibility. A simple policy lets your team use AI productively while protecting your customers and your business.

The one-page policy

1. Approved tools

List the AI tools your team may use for work, and which accounts. Business or team plans usually offer better data controls than free personal accounts, so check each tool's data settings before approving it.

2. Never paste in

• Customer personal data: names with phone numbers, addresses, ID numbers
• Financial details: bank details, card numbers, unpublished financials
• Passwords, API keys or login details
• Confidential contracts, pricing agreements or client documents
• Anything covered by an NDA

When in doubt, remove names and numbers first, or ask.

3. Always check before using

• Facts and figures: AI tools can state wrong information confidently. Verify anything you publish or send.
• Legal, tax, medical or compliance content: must be reviewed by a qualified person.
• Tone: make sure it sounds like your business, not a generic template.

4. Be honest with customers

Don't present AI-generated content as something it isn't, for example a fake testimonial or an invented case study. If a customer is chatting with an AI assistant, they should be told.

5. Ownership

The person who sends or publishes something is responsible for it, whether or not AI helped write it.

6. Questions and incidents

Name one person to ask about new tools, and to tell if something sensitive was pasted in by mistake. Mistakes reported quickly are much easier to handle.

Rolling it out

1. Share the policy in a short team meeting, with examples of what's fine and what isn't.
2. Ask everyone to confirm they've read it.
3. Review it every six months. The tools change fast.

Good uses to encourage

The point of a policy is to make AI use safer, not rarer. Encourage your team to use it for first drafts of emails and posts, summarising long documents, brainstorming ideas, and turning rough notes into clean formats, with a human check before anything goes out.