Should You Tell Customers When AI Wrote Your Marketing? A Disclosure Guide

09 September 2026·6 min read

Quick answer: Most AI-assisted marketing copy — blog posts, social captions, ad variations — doesn't need a formal disclosure if a human reviewed, fact-checked and approved it before it went out, the same way you wouldn't disclose that a copywriter or template was used. Disclosure becomes important when AI is generating something customers reasonably assume is human — like a review response, a personal testimonial, or a chatbot they think is a person — or when AI-generated imagery could mislead someone about what they'll actually receive. ✨

The honest answer to "should I disclose AI-generated marketing?" is: it depends what the customer would feel misled about if they found out, not whether AI was technically involved. Nobody expects a disclosure that a graphic designer used a template, or that a copywriter used a thesaurus. But if a customer reads a five-star review response and assumes the business owner personally wrote every word from the heart, and instead it's a slightly-edited AI draft sent to every reviewer with minor swaps — that's a different kind of gap. 💖 The test isn't "was AI involved," it's "would this specific customer, in this specific context, feel deceived if they knew."

What most businesses get wrong

Two opposite mistakes show up constantly. The first is businesses who slap "AI-assisted" disclaimers on everything out of anxiety, which is unnecessary for most day-to-day marketing content and can actually undermine trust by making customers wonder what else needs a warning label. The second, more common mistake is businesses who use AI to generate content that impersonates human judgement or human experience — a fake "our founder's story" blog post nobody at the business actually wrote or reviewed, AI-written testimonials, or a chat widget presented as "Sarah from our team" when Sarah doesn't exist. That second category is where real reputational and, in some cases, legal risk sits — particularly around misleading or deceptive conduct rules that apply regardless of whether a human or a tool produced the misleading claim.

The other overlooked issue: outsourcing the judgement entirely. If nobody on your team reads AI-generated copy before it publishes, you're not just risking a disclosure problem — you're risking factual errors, tone that doesn't sound like you, or claims your business can't actually back up. That's a bigger problem than disclosure.

The disclosure decision checklist

Run each piece of content through these questions before publishing:

1. Would a reasonable customer assume a specific human wrote or said this personally?
(e.g. a founder's personal reflection, a review reply signed by a named person) → Lean toward disclosure or, better, actually have that person review and personalise it.

2. Does this content represent something as real that isn't?
(AI-generated "before and after" images, fabricated case studies or testimonials, a chatbot presented as a named team member) → Disclosure or a rewrite is not optional here — this is a trust and, potentially, an accuracy-in-advertising issue.

3. Is this general marketing content (blog posts, ad copy, social captions, product descriptions) that a human has reviewed and stands behind?
→ No disclosure needed. Treat it like any other drafting tool.

4. Is a customer interacting with something live (chatbot, voice assistant) where they might reasonably think they're talking to a person?
→ Disclose plainly and early ('You're chatting with our AI assistant' beats letting someone find out three messages in).

If in doubt, ask: "If this customer found out afterwards, would they feel lied to, or would they shrug?" Shrug = no disclosure needed. Feel lied to = disclose or change the content.

A B2B firm using AI to draft blog posts: Every post is drafted with AI assistance, then edited, fact-checked and approved by the same person who'd have written it manually. No disclosure used — this is standard editorial practice, not deception, and matches how most publications already work with drafting tools.

A local service business using an AI chat widget on its website: The widget answers common questions after-hours. A small line at the top of the chat window reads "You're chatting with our automated assistant — for anything urgent, call [number]." Enquiries jump because people trust it more once they know what it is, not less.

A business tempted to use AI-generated 'customer' testimonials to fill a thin reviews page: This one's a hard no regardless of disclosure — fabricated testimonials, disclosed or not, risk misleading customers about real experiences with the business and can breach advertising standards. The fix isn't a disclaimer, it's collecting real reviews and, in the meantime, being upfront that the business is newer and building its review base.

How to word a disclosure that doesn't sound defensive

When disclosure is warranted, the tone matters as much as the fact of it. Avoid clinical, legalistic phrasing like "This content was generated using artificial intelligence and may not reflect the views of a human author" — it reads like a warning label and makes customers more suspicious, not less. Better: state it plainly, briefly, and move on. "Drafted with the help of AI, reviewed by our team" or "You're speaking with our automated assistant" both do the job in under ten words without sounding like a confession. Place it where it's relevant — at the top of a live chat, in a small footer note on a review-reply policy page — not as a giant banner on every blog post, which overcorrects into the anxious-disclaimer trap.

💡 The safest rule of thumb: disclose the interaction, not the tool. Customers generally don't care whether a spreadsheet or a search engine or an AI model touched a piece of content on its way to them. What they care about is whether they're talking to a real person when they think they are, and whether what they're shown is genuine. Anchor every disclosure decision to that, not to the technology itself.

Mistakes to avoid

Publishing AI-drafted content with nobody checking it for accuracy — a disclosure doesn't fix a factual error.
Using AI to generate fake reviews, testimonials or case studies, disclosed or not — this is a trust and compliance risk, not a wording problem.
Over-disclosing on routine content (blog posts, ad copy) in a way that makes customers more suspicious of everything you publish.
Letting a chatbot imply it's a named human team member without correcting the impression when asked directly.
Treating disclosure as a one-off decision — as AI tools get used in new parts of the business, revisit the checklist for each new use case.

Frequently asked questions

Is there a legal requirement in Australia to disclose AI-generated marketing content?

There's no single blanket law requiring an "AI-generated" label on marketing content, but general consumer law around misleading and deceptive conduct still applies regardless of how content was produced — so the risk sits in what's claimed, not in the drafting tool. Rules are evolving in this space, so check current guidance for your industry before finalising a policy.

What about AI-generated images in ads — do those need disclosure?

If the image could reasonably lead a customer to expect something different from what they'll receive (a venue, a product, a result), that's a real problem regardless of disclosure — the fix is not using misleading imagery, not adding a small-print warning under it.

Does using AI for internal drafting but having a human do the final edit still count as 'AI-generated' for disclosure purposes?

Generally no — once a human has reviewed, fact-checked and taken ownership of the content, it's reasonable to treat it the same as any other drafted content. The honest nuance here: 'reviewed' should mean genuinely read and checked, not a rubber stamp, because a poor edit is where accuracy problems slip through.

Will disclosing AI use hurt trust more than it helps?

It depends entirely on context and wording. Disclosed plainly in situations where it matters (a live chatbot, an automated response), it tends to build trust because customers feel informed rather than tricked. Disclosed anxiously on everything, it can make a business look uncertain about its own content — so calibrate to the checklist rather than defaulting to maximum disclosure.

Keep reading 🤍

Written by

Kate, founder of Chronically Online

I help Gold Coast and Brisbane businesses grow with branding, websites and marketing that actually works.

Work with me ✦