Loch Ness · HUMAN WRITING · Writing with AI
Part of the guide: Do AI Detectors Actually Work? What the Research Says
By Camille G., founder of Loch Ness · Published September 29, 2026 · Last updated September 29, 2026
Short answer: the line most people care about is who produced the words. Using AI to check spelling or look things up is one thing. Letting it write sentences that end up in your text is another. Different platforms and labels draw the line in slightly different places, so here’s a clear map of where it sits, and how to be honest about it.
Human Origin Label · For writers Wrote it yourself, with a little help? Let readers see where the line is. Show your process once, live or in a screen recording, and get a public page that says what was reviewed and how. Get verified, it’s free → Not ready for a review? Get the free self-declared badge →The distinction publishers use: generated vs assisted
Amazon’s Kindle Direct Publishing draws the clearest line. If an AI tool created the text, images or translation, that counts as AI-generated and must be disclosed, even if you edited it heavily afterwards. If you created the content and used AI to edit, refine, error-check or brainstorm, that counts as AI-assisted and doesn’t need to be disclosed (publishing.co.uk, Writers of the West).
In other words, it’s the starting point that counts. Writing a paragraph and asking AI to polish it is assisted. Asking AI for a paragraph and polishing it yourself is still generated.
What authors’ organizations accept
The Authors Guild, which runs the Human Authored certification for books, takes a similar line: authors can still qualify if they use AI for spell-checking or research, as long as the literary expression itself comes from the author (AP, via Yahoo News).
A practical map
| Use of AI | Amazon KDP | Human Origin charter |
|---|---|---|
| Spelling and grammar fixes | Assisted, no disclosure | Allowed |
| Research, looking things up | Assisted | Allowed |
| Brainstorming ideas, then writing yourself | Assisted | Allowed |
| AI outline used as the structure of the piece | Assisted | Not allowed |
| AI rewriting your sentences or paragraphs | Assisted if you wrote them first | Not allowed |
| AI drafting text you then edit | Generated, must disclose | Not allowed |
You’ll notice my label is stricter than Amazon. That’s on purpose. A disclosure rule decides what a platform needs to know; a human-made label makes a promise to readers, so it has to hold up when they take it literally. The full rules, discipline by discipline, are in the Human Origin Creator Charter.
How to be honest about it
- Say what you did, specifically. “I used a grammar checker” and “AI drafted the first version” are both fine things to say. Being vague is what erodes trust.
- Keep your process visible. Drafts and version history protect you if anyone questions your work later.
- Don’t let a tool smooth your voice away. Aggressive AI rewriting is also what makes texts read like everyone else’s.
If your work has already been questioned, start here: how to prove you wrote it. And if you want readers to see where you stand before anyone asks, here’s how human-written content labels compare.
Your words, your process, checked once and shown publicly. See how the Human Origin review works →Full disclosure: Human Origin, the label mentioned on this page, is something I created. I’m flagging it so you can weigh what I say about it. The facts above rest on the sources below.
Sources
- publishing.co.uk — KDP AI disclosure rules: what Amazon requires (and what it doesn’t)
- Writers of the West — Amazon KDP AI-generated content policy
- AP via Yahoo News — Authors Guild sets up “Human Authored” portal
Update log
September 29, 2026 — First publication.
