Yes, and it can be one sentence at the bottom of the page. No law requires it. The reason to add one is simpler than compliance… you would rather your reader hear it from you than work it out on their own.
What it actually is
A short line at the end of a blog post or newsletter saying how the piece was made. Not a legal page. Not a policy document. A footnote.
Mine reads:
How this was written: The observations are mine, from years of experience, client work, and platform watching. AI helps me draft, tighten, and check my grammar. It doesn’t decide what’s worth saying. AI disclosure v1.0 … July 2026.
Three sentences and a version number. It took longer to decide to publish it than to write it.
Why bother when nobody is asking
Most readers will never ask how your content was made. The ones who do tend to be the ones deciding whether to hire you.
Picture the moment you are actually protecting against. A potential client is reading your work, likes it, and somewhere in there decides on their own that a machine wrote it. Nothing was hidden from them, exactly. You just never said. Your thinking quietly turns into output in their head, and you are not in the room to explain the difference.
A line at the bottom closes that gap before it opens. It costs you nothing when the reader does not care, and it saves you the whole thing when they do.
Why this is coming up now
Labeling stopped being a debate and started being infrastructure. Four things happened while you, the business owner was busy running a business.
- Substack turned on AI detection through a partnership with Pangram, so a reader can check an article themselves.
- LinkedIn has been reading Content Credentials since 2024. Images carrying that data show a small CR badge, and clicking it reveals which tool made the image. This can happen even in Canva editing and filters.
- Spotify launched AI Credits in April 2026, surfacing disclosed AI involvement inside song credits.
- Anthropic confirmed in August 2026 that its newer Claude models embed an invisible watermark in generated text.
None of this establishes who wrote anything. Editing your own paragraph with AI leaves the same trace as generating one from scratch, and heavy rewriting removes the trace entirely. The marks show a tool was involved, nothing more.
What it does mean is that the question reaches your reader from more directions than it used to. You want your answer already sitting at the bottom of the page when it does.
What people are actually reacting to
Not AI. Nobody has ever been bothered by an AI-assisted FAQ page or a tidied-up service description.
What lands badly is discovering the personal parts were not personal. The founder story. The client results. The paragraph where you explained why you do the work this way. Those pieces carry the relationship, and readers feel differently about them the moment they suspect nobody was there.
Which is useful information, because it tells you where to keep your hands on the work. It also makes the footnote easy to write honestly, since you already know which parts are yours.
What to put in it
Not legal advice – just a suggestion on what to write. Three things. One sentence each is plenty.
Where the thinking came from. Your experience, your client work, your observations.
What the tool did. Drafting, editing, grammar, formatting… whatever is true for you.
What the tool did not decide. Usually this is what to say and whether it was worth saying.
Include the small tools that do not feel like AI. Grammarly counts. A disclosure that quietly skips the comfortable parts is not really a disclosure, and the omission is the thing that would embarrass you later.
Add a version number
Put v1.0 and a date at the end of the line.
This is a small thing that does a lot of work. A dated line reads like a practice. An undated one reads like a page somebody wrote once and forgot. When your workflow changes, and it will, you change the number instead of quietly editing the old wording.
Where it goes
Bottom of blog posts and newsletter issues. Same wording every time, so regular readers stop noticing it and new ones can find it when they go looking.
You do not need a separate page for this to be worth doing. A consistent footnote is the whole first step.
The takeaway
Write the sentence. Put it at the bottom of your next post. Keep the wording the same until your process actually changes.
The advantage is entirely in the timing. Said on your own schedule, it reads as how you work. Said after someone asks, the same words read as a defensive explanation.
Frequently Asked Questions
Do I legally have to disclose that I used AI to write content?
No. At the time of publishing, we could find no general legal requirement for a business publishing its own marketing content in the US. The case for disclosing is about trust and timing, not compliance.
What should an AI disclosure line say
Where the thinking came from, what the tool did, and what the tool did not decide. Three sentences is enough. Add a version number and a date so it reads as a current practice.
Does using AI for grammar and editing count as AI use?
Yes, and it belongs in the line. Tools like Grammarly are AI. Leaving out the parts that feel harmless is what makes a disclosure look incomplete if anyone examines it.
Will disclosing AI use make readers trust me less?
Readers react to substitution, not to tools. The problem shows up when someone suspects the personal parts of your content were not personal. AI slop and generic content also come into play – we write more about that in another post. Saying plainly which parts are yours and what you use AI for, addresses that directly.
How this was written: The observations are mine, from years of experience, client work, and platform watching. AI helps me draft, tighten, and check my grammar. It doesn’t decide what’s worth saying. AI disclosure v1.0 … July 2026.
