AI Use Policy
Considered · AI governance + training
A practical AI acceptable-use policy for your team — what tools are approved, what data can never be pasted in, who reviews edge cases — drafted in plain English without enterprise-suite ceremony.
The included items, listed plainly. No asterisks, no fine print.
- 01Drafted AI acceptable-use policy, tailored to your business
- 02A short, named list of banned data types
- 03Named escalation paths for uncertain cases
- 04Half-day team training, recorded for new starters
- 05Annual review and refresh
Three steps, then quiet.
- Step 01
Conversation, then policy
We start with a one-hour conversation about which AI tools your team is already using (and which they would like to use). The policy is drafted around that reality — not a generic template, but a document your team actually recognises.
- Step 02
Banned-data list and review path
A short, named list of data types that should never be pasted into an AI tool — customer health records, payment details, anything covered by NDA. A named person who reviews uncertain cases. The whole list fits on one page.
- Step 03
Half-day training, recorded
We run a short training with your team — the policy in five minutes, then the rest in worked examples. The session is recorded so new starters get the same induction. Annual refresh as the AI tooling landscape moves.
Built for these operators.
- Small teams that have started using ChatGPT, Copilot, Gemini, or Claude and noticed they have no policy
- Boards and owners required to demonstrate AI governance for insurance, tenders, or audit
- Operators who want their staff to keep using AI — productively, but with a few clear rules
- Schools and nonprofits needing a defensible position on student / member data and AI tools
And what it is not.
- An enterprise compliance suite. There is no purchasing-portal, no AI-risk-register dashboard, no audit-log SaaS. The policy is a document.
- A blocker. We start from "AI tools are useful and you should keep using them" — the policy says how, not whether.
- AI consulting. We do not pick which AI tool you should adopt or build prompts for you. The policy works whichever tool you choose.
The questions we hear frequently.
If yours is not here, write to us. We are honest when the answer is "it depends".
- 01
Why do we need a policy at all?
Three reasons, in order of urgency. First, to stop sensitive data leaking into a hosted AI tool by accident. Second, to give your insurer and your customers a defensible answer when they ask. Third, because the EU AI Act and Australian equivalents are moving, and a written policy is cheaper to update than a culture is to change.
- 02
How is this different from a privacy policy?
A privacy policy is for your customers. This is for your staff. The privacy policy says how you handle data; the AI use policy says how your team is allowed to use AI tools when they handle data.
- 03
Can you write it for a specific tool — Copilot, ChatGPT, Claude?
Yes. We tailor the approved-tools list to whatever you actually use. The policy stays meaningful even when you switch tools.
- 04
What if our staff are already using AI tools the policy would ban?
Common, and the conversation we usually have on day one. The policy is written in service of getting that usage to a defensible place — sometimes by approving a tool with conditions, sometimes by switching to a safer alternative. Always with the team, not at them.
Curious about AI Use Policy?
Drop us a line. We will set up a friendly, no-pressure conversation and walk you through whether this is a fit for your situation.