AI Transparency Statement
Last updated: 14 September 2026
We know “AI” makes a lot of authors nervous — and honestly, it should make you ask questions. So here’s exactly what AuthorOrbit’s AI does, what it doesn’t do, and why your work stays yours.
The short version: this is AI used exactly as it should be. Not to write for you, not to think for you — but to take the admin off your plate so you get more time back for the thing you actually do best. Writing.
It only ever works with your own material
AuthorOrbit’s AI features are built to read and organise your writing, not anyone else’s. When it drafts a caption, summarises a royalty statement, or sketches a Plot Bible, it works only with the manuscript, notes and details you’ve put into your own account. It isn’t trained on other authors’ manuscripts, doesn’t pull from a shared pool of author data, and doesn’t use your material to help anyone else’s book. Your material stays exactly that: yours.
Where your data actually lives
Everything you enter into AuthorOrbit — your manuscripts, your royalty figures, your contacts, your notes — lives in a database on your own computer. It isn’t uploaded to a server, backed up to a company’s cloud, or visible to us, to Anthropic, or to anyone else. If you never touch an AI feature, nothing you write ever leaves your machine.
The one exception is deliberate and visible: when you use an AI feature — asking Claude to sketch a Plot Bible from your chapters, or draft a caption — the specific content involved in that request (not your whole library, just what’s relevant) is sent securely to Anthropic’s Claude API to generate the response, then returned to you. Under Anthropic’s current commercial terms, that content isn’t used to train any model and isn’t retained beyond a short operational window. But it’s honest to say: at the moment you use an AI feature, that piece of content briefly leaves your machine to be processed. The rest of AuthorOrbit — your whole library, your admin, your records — doesn’t.
It doesn’t write your book, or your ideas, for you
This isn’t generative fiction. AuthorOrbit’s AI won’t invent your plot, put words in your characters’ mouths, or draft a scene from scratch. What it does do is reflect back what you’ve already written: a Plot Bible or Character Bible sketched from your own chapters, a read on where your story’s tension is rising or sagging, a royalty statement turned into a tidy table. Every output starts with something you wrote or provided — it never starts with a blank page pretending to be your imagination.
Your voice stays your voice
The one place AI does actively “write” anything is social captions — and even there, it’s translating, not inventing. Tell it what a post is about (“cover reveal for Book 3”) and it drafts the caption in your voice, using the tone and style you’ve already established, for whichever platform you’re posting to. It isn’t choosing what to say about your book or making claims on your behalf — it’s putting your message into a ready-to-post shape, faster than doing it from scratch. You always see the draft before anything goes out, and the buy links it inserts point only to your own books.
No AI-generated images. Full stop.
AuthorOrbit will never generate an AI image — not for a cover, not for social media, not anywhere. That isn’t a setting you can toggle on: it’s a restriction written directly into the app’s code. If you want an image in a post, it has to be one you supply — your cover art, your photography, your own design.
About the Anthropic lawsuit you may have read about
You may already know that Anthropic, the company behind Claude, settled a major lawsuit in 2025 — $1.5 billion, one of the largest copyright settlements in US history — over pirated books used to help build its earliest models. We’re not going to pretend that didn’t happen, because it did, and it’s a fair thing to be angry about. As part of the settlement, Anthropic is required to destroy those pirated datasets and any copies derived from them.
What matters for you is what governs AuthorOrbit today, not what happened years ago while the underlying model was first being built. Under Anthropic’s current commercial terms, content processed through the API — including anything AuthorOrbit sends on your behalf — isn’t used to train Claude or any other model, and isn’t pooled with any other author’s data. That’s a current, contractual policy, not a promise resting on goodwill, and it’s a separate matter from the historical lawsuit. We think you deserve to know both things, not just the reassuring one.
Why this matters to us
As an author myself, I built AuthorOrbit because I wanted the admin off my plate — not my voice, my ideas, or my integrity. Every AI feature in this app is built to save you time on the parts of the job that were never really “the writing” in the first place, and to leave everything that makes your work yours exactly where it belongs: with you.
That’s the whole point of AI here: productivity, not creativity. A tool that clears space in your day, not one that fills the page for you — so you can spend more of your time doing what you do best.
Questions
Questions about how this works? Email contact@authororbit.co.uk.