Your AI Use Policy: Why People Matter More Than AI

What is Microsoft's humanist AI code of conduct?

Microsoft published a 37-page "humanist AI code of conduct" built on one core line: people matter more than AI. It argues that AI models aren't conscious and shouldn't pretend to be. For a small business owner, that document isn't just corporate philosophy — it's a free template you can adapt into your own AI use policy.

If you're rolling out AI agents for customer support, marketing or admin work, a short internal policy keeps everyone aligned. You don't need 37 pages. You need clarity on what AI does, what people still decide, and how you'll handle mistakes.

Why does an AI use policy matter for a small business?

It matters because AI is now touching customer conversations, invoices and inboxes — often without a written rule about who's responsible. An AI use policy sets expectations before something goes sideways.

A policy also builds trust. Customers in Vancouver, across British Columbia and anywhere in Canada increasingly want to know when they're talking to a bot. Staff want to know what they're allowed to automate. A one-page document answers both.

What should your AI use policy actually say?

Start with the principle, then get specific. The humanist framing gives you a simple spine: people first, AI as a tool.

Here's a practical outline you can copy:

• Purpose. State plainly that AI supports your team — it doesn't replace judgment.
• Scope. List which tools you use (for example, an AI agent handling first-line support or drafting marketing copy).
• Human oversight. Name who reviews AI output before it reaches a customer or a contract.
• Transparency. Explain when customers are told they're interacting with AI, and how they can reach a person.
• No fake humanity. AI shouldn't claim to be human or simulate emotions it doesn't have.
• Data handling. Say what information AI tools can and can't access.
• Escalation. Define what triggers a handoff to a human, and how fast.
• Review. Set a date to revisit the policy — AI changes fast.

Keep the language plain. If a new hire can't understand it in five minutes, it's too long.

How do you apply the policy to AI agents in practice?

You apply it by building the rules into the agent itself, not just the document. A policy on a shelf changes nothing; a policy wired into your workflows changes everything.

Set clear boundaries for each AI agent

Give every agent a job description. A support agent answers common questions and books appointments. It does not negotiate refunds. A marketing agent drafts posts; a human approves them. Boundaries prevent the awkward moments that erode customer trust.

Keep a human in the loop where it counts

Not every task needs review — that would defeat the point of automation. But complaints, pricing exceptions and anything legal should always land with a person. Set that threshold once and let the agent route accordingly.

Be honest about the handoff

When a customer asks to speak to someone, the agent should connect them quickly and say so clearly. No pretending, no loops. This is the practical version of "models aren't conscious and shouldn't imitate it."

Review results monthly

Pick a few signals: response times, escalation rates, customer feedback. If the agent is creating more work than it saves, adjust the scope. A policy is a living document.

What does this look like in day-to-day operations?

Picture a small clinic or trades business using an AI agent to answer after-hours messages. The agent books appointments and answers FAQs. Anything about billing or a complaint goes to a human the next morning. Customers are told up front they're chatting with an assistant. Staff know they own the final call.

That's it. No drama, no over-engineering — just a clear line between what the AI handles and what people decide. [your real example here]

The payoff is real: faster replies, fewer missed leads and a team that isn't afraid of the tools.

Start with one page, not thirty-seven

You don't need to mirror Microsoft's whole document. Borrow the principle — people matter more than AI — and write one page that fits your business. Then make sure your AI agents actually follow it.

Want help mapping your operations and building an AI use policy that matches how you work? Book a free AI audit and we'll walk through it together.

FAQ

Do I legally need an AI use policy?

In most cases there's no specific law requiring one, but privacy and consumer protection rules still apply to how you handle data and customer interactions. A written policy helps you stay compliant and consistent.

Can I just copy Microsoft's code of conduct?

You can use it as a reference, but it's written for a large enterprise. A small business needs something shorter and more specific to its tools, customers and team.

How long should an AI use policy be?

One to two pages is plenty for most small businesses. Focus on scope, oversight, transparency and escalation — the parts that actually affect daily work.

Who should write it?

The owner or a manager who understands the workflows, ideally with input from the people using the AI tools every day. Review it with anyone handling customer data.