# Write a simple AI use policy for a small team

Original 5HO practice material. Fictional example. Reviewed 5 October 2026.

## Useful result

You'll finish with a short policy people can apply to daily work.

A short team rulebook beats a vague “use AI responsibly” message. Write down what people can use, what they can share and who checks the result. Keep it readable by the people doing the work.

## Set up

Allow 30 minutes for a draft. Bring your organisation’s existing rules and a decision owner. This practice document does not replace legal, security or employment advice.

## Practice input

Fictional team: five staff. Approved account: company AI workspace. Allowed example: public brochure drafts. Excluded until approval: customer records and contracts. Jo owns policy questions. Outputs must be checked before external use.

## Do the task

1. Separate approved facts from policy decisions still awaiting agreement.
2. Draft one page with allowed work, exclusions, review and incident contact.
3. Have the responsible people approve it before calling it a team policy.

## Prompt

Draft a plain-English AI-use guide from these facts. Include approved account, allowed inputs, excluded inputs, review responsibility and questions for Jo. Mark missing decisions as Draft: needs approval.

## Refine

Add a practical example of permitted brochure drafting and an example that must wait. Avoid claiming this policy guarantees compliance.

## Expected checks

The result should name an actual decision owner and distinguish draft guidance from an approved rule. Missing data-handling decisions must not be filled in by the model.

## Save

Save policy-draft.md with an owner and review date. Publish it internally only after the required approval.

## Sources

- Connected apps and access: https://help.openai.com/en/articles/11487775-connected-apps-in-chatgpt