What happens if a client finds out I used AI without telling them?

Research on agency-client relationships shows clients who feel misled about AI use rarely confront the freelancer directly. Most quietly move the work elsewhere and stop referring them, without ever raising it as a complaint. The risk isn't a conversation, it's silent churn.

← DisclosureTrail For freelancers & agencies

Should I Tell My Client I Used ChatGPT?

Q: Should I tell my client I used ChatGPT?

In most cases, yes, if AI did meaningful drafting rather than a small assist like spellcheck. It's rarely a strict legal requirement, but client contracts increasingly expect it, and clients who find out on their own tend to quietly stop hiring you rather than complain. A short factual mention costs you almost nothing; getting caught costs the relationship.

You got it done fast with AI and now you're staring at the send button, wondering if you need to mention it. Short answer, the real risk if you don't, and exact wording if you do.

Client already found out and is accusing you now, not just weighing whether to mention it? Get the evidence checklist and a ready-to-send response for that exact conversation. Get your evidence packet →

The honest answer Yes, in most cases — if AI did meaningful drafting rather than a light assist like spellcheck or a grammar pass. It's rarely a hard legal requirement for a single freelancer (see the full legal breakdown), but that's the wrong question to be asking. The real question is what happens to the relationship if they find out on their own instead of hearing it from you.

What actually happens if you don't say anything

Not a confrontation. That's the part that catches people off guard. Clients who feel misled about AI use rarely bring it up directly — a Fractl survey found 84% of clients want AI content labeled, but most don't ask outright if it wasn't. What happens instead: the next project quietly goes to someone else, and the referrals stop, with no conversation and no chance to explain. You don't get told you're being replaced for this reason. You just notice the work drying up a few months later.

Compare that to the cost of mentioning it upfront: a sentence. That asymmetry — small cost now, quiet and permanent cost later — is the actual reason to disclose, more than any specific law.

When it matters most

Situation	Say something?
ChatGPT drafted most of the copy, you edited it	Yes — a quick line, low effort either way
Client's contract says "original work," AI wasn't discussed	Yes — silence isn't the same as permission
Deliverable goes out under the client's name to their audience	Yes — it's their exposure if it comes up later, not just yours
You used AI to brainstorm, then wrote it yourself	Optional — but a one-line record costs nothing

Exact wording, if you want it short

You don't need a legal paragraph. A factual sentence in the delivery email or a note on the invoice does the job:

"This draft was AI-assisted using [tool] and reviewed and edited by [your name] before delivery on [date]."

That one line does three things: names the tool, names the human who's accountable for the result, and puts a date on it. If a client ever asks later, you have a record instead of a memory.

Why this beats staying quiet Saying it first means you control how it's framed — a professional using a modern tool well, not someone hiding something that got noticed later. The framing only works if you're the one who says it.

If you want more than one line

For higher-stakes work — a full campaign, a client who's explicitly asked before, anything published under their name — a proper dated statement holds up better than a sentence in an email thread. See what a full disclosure statement should contain, with a filled-out example.

Generate the sentence, or the full statement

DisclosureTrail fills in tool, reviewer, date, and involvement level, and gives you a downloadable record with a reference ID. Free, no login, nothing stored on our end.

Generate a disclosure statement →