This lesson is being recorded
Every video of this course is being uploaded to my YouTube channel. Subscribe so you do not miss it.
Bruno Ashford @brunoashfordSubscribeModule 11 · Lesson 11.6 · 9 min
Your AI Use Policy
Write a one-page internal AI use policy that your team can actually follow.
Use ← → to move. The marker highlights text and the pencil draws on the slide; navigation locks while you annotate. Undo removes the last mark, Esc exits.
1Start from the template
Copy the one-page AI Use Policy template from this module's materials into your business folder. It has eight short sections and fits on one printed page.
2Let Claude draft it from your real setup
Ask Claude to fill in the template using your settings, your approval matrix and your privacy rules, so the policy describes what is actually enforced rather than wishful thinking.
Fill in materials/11-ai-use-policy-template.md for our business using what is actually configured: the permission rules in .claude/settings.json, the approval matrix, the Privacy section of CLAUDE.md and our list of approved tools. Keep it to one page, in plain English a new employee understands. Mark anything you had to assume with ASSUMPTION.3Review it as the owner
Read every line and ask: would I be comfortable if a client read this? Can a new hire follow it without asking me? Remove anything vague. For regulated sectors (health, legal, finance), have your lawyer or professional body review it.
4Run the security checklist
Go through the Agent Security Checklist for every agent, skill and routine you have built so far. Fix what fails before you give more autonomy.
Go through materials/11-agent-security-checklist.md for this project. For each item, tell me pass, fail or not applicable, with the evidence (file and setting). List the failures first, with the fix for each.5Share it and review it quarterly
Share the policy with the team, ask each person to confirm they have read it, and put a quarterly review in the calendar. AI tools change every few months; your policy should change with them.
Your team is already using AI. The only question is whether they are using it with rules you chose or rules they made up. A policy is not bureaucracy. It is the thing that lets you say yes to AI with confidence, and it is the first document a larger client, an insurer or a partner will ask for.
Most AI policies fail because they are twelve pages of legal language nobody reads. Yours will be one page: what is allowed, what is not, what data can be used, what needs approval, and who to ask. You will draft it with Claude from everything you set up in this module, then make it yours.
Key points
- Your team already uses AI; the policy decides whose rules they follow.
- One page that people read beats twelve pages nobody does.
- Describe what is actually enforced, not what you wish.
- Review it every quarter as the tools change.
Avoid
- Copying a generic corporate policy that does not match how you work.
- Writing rules that are not backed by any setting or approval step.
- Never telling the team the policy exists.
- Treating the policy as finished forever.
A 25-person roofing company had office staff using three different AI tools with no rules, and one estimator had pasted a customer's insurance claim into a free chatbot. The owner drafted a one-page policy with Claude in under an hour: approved tools only, no customer documents in personal accounts, approval for anything sent to insurers, and the office manager as the contact for questions. When a commercial client later asked for their AI policy during a tender, they sent it the same day.