Wisdom & Insights

Should I Tell Clients I Use AI? Write Your AI-Use Policy First

Smiling woman in glasses and a mustard shirt, chin on her hands, at her desk

You used AI this week. Maybe it tidied your call notes, sketched an outline or turned a rambling voice memo into an email you could actually send. And somewhere in the back of your mind, a little question keeps tapping: should I tell clients I use AI, or is it better not to bring it up?

It usually comes with a knot in the stomach. Will they think you’re cutting corners? Or, if you stay quiet and they find out later, will they feel tricked?

Here’s what 20+ years of client work has taught me: the slightly awkward conversation up front is almost always kinder than the one after someone feels surprised.

Short answer: Yes. Tell clients you use AI in plain words, before they ask, and back it up with a short written AI-use policy. The policy covers what you use AI for, what you never use it for, how you protect client information, who reviews the work, who owns the deliverables and how clients can opt out. Put it where clients actually read it: your services page, proposals, contract and onboarding.

Should I tell clients I use AI tools, or only if they ask?

Tell them before they ask. Waiting for the question feels safer, but it hands your client the job of finding out.

That’s not just my opinion. An IBM Research study on AI disclosure in freelance work used interviews and surveys of freelancers and clients and found an expectation gap. Freelancers often mention AI only when asked, because they assume clients can already tell. Clients are far less confident that they can spot AI-assisted work, and they prefer to be told up front.

So the silence you meant as polite can read as hiding. (Wild, right? Two reasonable people, completely missing each other.)

The same study found that unclear or missing client AI policies left freelancers misreading what clients expected, and the researchers point to a need for clearer guidelines. Your client’s policy isn’t yours to write. Yours is, and it ends a lot of guessing on both sides. So go first.

Should I tell clients I use AI to write?

Yes, and be specific, because “I use AI” can mean anything from fixing a typo to writing the whole thing. Writing deserves extra clarity, since clients are often paying for your voice and thinking. Name the role AI plays:

  • Thinking partner: angles, outlines or headline options you then choose from.
  • Editor: tightening sentences, catching typos, checking a draft against the brief.
  • First drafter: a rough version from your notes that you rewrite in your own voice.
  • Never: invented quotes, stories, statistics or unchecked facts.

“I use AI as an editor and a brainstorming buddy, and the ideas, interviews and final words are mine” is a sentence most clients will be glad to read. If you use AI as a first drafter, say that too. Clients with strict rules of their own deserve to know before they hire you.

A woman in a checked shirt holding a pencil to her lips and gazing into the distance at a white desk with a laptop and coffee, asking herself: should I tell clients I use AI?

What should an AI-use policy for clients include?

A one-page AI-use policy covers six things: what you use AI for, what you never use it for, how you protect client information, who reviews the work, who owns it and how clients can opt out. I call it the Open Studio Policy: like an artist’s open studio, people can walk in and see how the work gets made.

SectionThe client’s questionWhat to write
1. What I use AI forWhere does AI show up in your work?Specific tasks: research, outlines, first drafts, code, meeting notes
2. What I never use AI forWhere are your lines?Strategy decisions, unchecked facts, talking to clients in your name
3. How I protect your informationIs my stuff going into a chatbot?What you never paste in, which tool settings you use, how you remove private details
4. Who reviews everythingWho’s accountable?You review, edit and fact-check every deliverable before it reaches them
5. Who owns the deliverablesIs the finished work mine?Final deliverables are theirs as your contract describes, with exceptions (licensed fonts, stock photos) named
6. How to opt outWhat if I don’t want AI on my project?How to ask, by when, and whether it changes the timeline or price

Keep it to one page. If your policy needs a scroll bar and a glossary, it’s no longer a policy. It’s a hiding place.

Where do I put my AI policy so clients see it?

Put a version of it everywhere a client makes a decision. Once it’s written, “how do I tell clients I use AI?” stops being a question you answer on the spot. It’s the same message in four sizes:

  • Services page: two or three sentences (“Yes, I use AI tools, carefully. Here’s exactly how.”) plus a link to the full policy, so clients find out before they book a call.
  • Proposals: the two-line script below, right next to your process or timeline.
  • Contract: a clause on AI use, confidentiality and ownership that matches your policy. Ownership of AI-generated material is still being worked out legally, and I’m not a lawyer, so have one review the exact wording.
  • Onboarding: the full policy in your welcome packet, plus one intake question: “Do you have an AI policy of your own, or anything you’d like me to avoid?”

That last question matters, because some clients have rules from an employer, a board or an industry.

What do I say about AI in a proposal?

Two sentences are enough. Here’s a script to adapt:

I use AI tools for research, outlining and editing, and I personally review and refine everything before it reaches you. Your confidential information stays out of AI tools unless you say otherwise, and if you’d prefer I skip AI on your project, just let me know.

It doesn’t apologize or over-explain. It sounds like a professional describing her process, because it is.

What does a simple AI-use policy look like?

It’s six short sections a client can read in a minute. Here’s one for a copywriter to adapt:

  • What I use AI for. Research summaries, outlines, headline options and editing.
  • What I never use it for. Strategy decisions, interview quotes, or anything presented as fact before I’ve checked it myself.
  • Your information. I don’t paste contracts, financial details or personal information into AI tools. I remove names where I can and turn off model training on my inputs wherever a tool allows it.
  • Human review. I personally review, edit and fact-check every deliverable. If my name is on it, I stand behind it.
  • Ownership. Final deliverables belong to you as described in our contract. Licensed elements, like stock photos or fonts, are listed with their terms.
  • Opting out. Prefer no AI on your project? Tell me before we start and I’ll work without it. If that changes the timeline or price, I’ll say so first.

What does honest AI disclosure look like in practice?

It names what AI did, what you decided and what a human checked. I’ll go first, since I’m asking you to. My Purposepreneur Toolkit, the web tools inside my own WordPress site, was built with AI (Claude, specifically). My disclosure, Open Studio style:

  • What AI did: helped me build it, fast.
  • What I did: made the calls on what the tools do and don’t do.
  • What got checked before launch: every calculation against an answer key, the US tax figures against the IRS and Social Security Administration, and the security.
  • Where you’ll meet AI: optional, capped AI drafting in two paid tools. Skip it if you like.

You can read the full version, which covers all six parts, on my AI Use Policy page.

And here’s the line I’ll always say out loud: my own builds go lightning fast, because I’m my own client. Client projects take longer. AI speeds up parts of the making. It doesn’t speed up the listening.

Put your version of that in your policy, so nobody hears “AI” and pictures a finished website by Tuesday.

Try this today: draft your Open Studio Policy (20 minutes)

  1. List every place AI touched your work in the last month. Tiny tasks count.
  2. Circle the tasks you’d happily name in public. If there’s one you’d rather hide, that’s information: change how you use it, or get comfortable saying it.
  3. Write your “never” list: three things AI doesn’t do in your business.
  4. Open the privacy or data settings in each AI tool you use. Can you keep your chats out of model training? Note what you can honestly promise clients.
  5. Fill in the six sections with a sentence or two each, then paste the script into your proposal template.

Then ask one trusted colleague, “Does anything here surprise you?” Whatever does is the line to rewrite.

If you want AI to sound like you before it touches client work

Your policy promises a human reviews everything. That’s much easier when you know exactly what “on brand” means. Brand North Star turns guided questions into a one-page brand brief, including your ideal client in their own words and voice guidelines you can hand to any AI tool. It’s $47, with AI help (its page explains exactly what the AI does). Then your policy can add: “Everything AI drafts is checked against my brand brief.”

Want a thinking partner for where AI fits in your client work? Bring it to coaching.

Quick answers

Do I have to tell clients I use AI?

Whether you’re legally required to tell clients you use AI depends on your contracts, your industry and where you work, so check your situation with a lawyer. Practically, telling clients you use AI is the safer default. Research on freelance work found clients prefer hearing it up front, and a short written policy prevents misunderstandings before they start.

What is an AI-use policy for freelancers?

An AI-use policy for freelancers is a short, plain-language page that tells clients how you use AI in your work. It covers what you use AI for, what you never use it for, how you protect client information, who reviews deliverables, who owns them and how clients can opt out. One page is plenty.

What if a client doesn’t want me to use AI?

If a client doesn’t want you to use AI, respect that and work without it on their project. Your AI policy should explain how to opt out, by when, and whether it changes the timeline or price. Ask about AI preferences during onboarding, so you learn this on day one rather than at the final review.

Is it unprofessional to use AI for client work?

Using AI for client work isn’t unprofessional when you’re honest about it, protect client information and personally review everything you deliver. What feels unprofessional to clients is surprise: finding out later, or receiving work nobody checked. Clear disclosure and human review make AI one more tool in a professional process.

You don’t have to hide your tools to be trusted. Open the studio doors, say it plainly, and let the right clients see how beautifully you work.

Leave a reply

Your email stays private. Required fields are marked *