Free tool, no email gate

Somebody on your team used AI this week. These are the rules they followed.

Every company already has an AI policy. In most of them it is unwritten, invented one shortcut at a time, and nobody knows what it says until something goes wrong. Answer eleven questions and this page writes the version you can actually hand out: your decisions, your language, ready to sign.

Make the calls

Eleven decisions, and the policy writes itself

AI Use Policy GeneratorRuns on your device

Step 1. Who this is for

The defaults below are the settings most owners of growing businesses land on once they have thought it through. Change the ones you disagree with and watch the document rewrite itself.

Company name Optional. Leave it blank and the policy reads as a template you can fill in later. What kind of work you do This decides which kinds of information the policy names as restricted.

Step 2. The two rules that actually matter

Almost every real incident traces back to one of these two. Get them right and the rest of the document is housekeeping.

May staff use public chat tools at all? Banning them outright is a real choice, and the policy says so plainly rather than pretending it never comes up. May customer information go into them? This is the line that costs money when it is crossed. The default is the safe one.

Step 3. What leaves the building

A draft nobody reads is the second most common way this goes wrong, and it is the one your customers actually see.

May client work be AI drafted? Must a person review it before it ships? Whoever signs off owns every fact in it. That sentence is the whole rule. Who approves a new AI tool?

Step 4. Meetings, disclosure, and what happens when it goes wrong

The awkward parts. Deciding them now is cheaper than deciding them in the middle of an incident.

Are AI notetakers allowed in meetings? If you allow them, the policy hands your team the exact sentence to say before recording starts. How is AI written text disclosed? What happens when someone breaks it? How often is the policy reviewed?

Step 5. Your policy

Complete and signable as it stands. Read it once with your own team in mind, change the sentences that do not sound like you, and put it in front of people.

Runs in your browser. Nothing you type is stored or sent.

A plain-language starting point built from your own answers. It is not legal advice and not a substitute for counsel. If your work is regulated, or your contracts dictate how customer information is handled, have your attorney read it before you circulate it.

Get the AI readiness audit Book a call

Want the rollout memo and the customer-data rules alongside it? They live in the Ground Rules Pack. Every self-serve tool sits in the toolkit.

The honest part

A policy is not the point. Being able to answer is.

The document this page builds is worth exactly as much as the conversation you have when you hand it out. Nobody reads a policy twice. What people remember is the ten minutes where the owner said, out loud, here is what you may do, here is the one thing you may never do, and here is who to tell when you get it wrong. Print the page, read it to the team, take the questions, and the rules survive contact with a busy Tuesday.

What this page cannot see is your actual exposure: which systems hold customer records, which contracts promise a client that their information stays inside your walls, and which jobs are quietly being done by a tool nobody approved. That is a different exercise, and the readiness audit is where it happens, on your real systems, with a person doing the looking.

Do we really need a written AI policy?

If nobody has written one, the team has already written it for you, one habit at a time. A page of plain rules costs an afternoon and settles the two questions that actually cause damage: what may be pasted into an outside tool, and who reads the work before a customer does.

Is this a legal document?

No. It is a plain-language starting point built from your own choices, not legal advice and not a substitute for counsel. If your work is regulated, or you carry contracts that dictate how customer information is handled, have your attorney read it before you circulate it.

What happens to what I type here?

Nothing. The policy is assembled by code running inside your own browser. There is no account, no upload, and no copy kept anywhere, which is why you can put your real company name in it.

Our team is fifteen people. Is this overkill?

Fifteen people is exactly the size where this bites. Big companies have a compliance function that says no for them. Growing businesses have a group of capable adults improvising, which works beautifully until the day one of them pastes a customer file into an outside tool because it was the fastest way to finish before pickup time.

AI Use Policy Generator

Policy written. Now the harder question.

Thirty minutes, free, in plain English. Bring the policy you just generated and one job you wish the team could stop doing by hand. We will tell you which one is worth automating first and what it would take.

Book the free 30-minute call

Want the ranked list instead? The AI readiness audit.

Next free tool, same no-gate rule: the two-minute readiness check. Whether your shop is ready to put AI to work at all.

hello@clariqadvisory.com