AI Ethics & Transparency Statement

1 · What our AI does, and what it does not

Write in a Click is AI-native, and we say so plainly. What changed in September 2026 is not whether we use AI but what we use it for.

Robin, our assistant, never changes your manuscript on its own. It reads your whole book — the manuscript, your Story Hub and your voice profile — and answers in a chat panel beside the page. Chat opens in Robin Mode, which is advice only. In Edit Mode, which you switch on, Robin can suggest changes; each one shows in the text for you to keep or reject, and nothing lands unless you keep it. Proofing runs when you ask it to, and changes only what you accept. Paid plans include optional tools that can draft a starting outline or cover art; nothing from them enters your chapters unless you choose it.

The Proofing Pass changes text only on request. It is a pass you start, not a layer that runs as you type, and it applies only the corrections you accept, one at a time.

2 · What we use AI for instead

We spend AI on work that was previously impossible, slow or expensive for an author to obtain — not on producing prose:

• Reader Pulse — predicted reader engagement and drop-off risk for every chapter, before a human reads the book.
• Draft from Book — world-building your chapters already established, surfaced for your approval when you run it from the Story Hub. A manuscript imported during onboarding is filed into the Story Hub directly.
• Writing voice analysis — what your prose sounds like, across five pillars.
• Writing Insights — the conditions under which you actually write well.

A full description of each, and of the plans they are available on, is on what AI does here.

3 · Your work, and your ownership of it

You own 100% of what you write. Robin never changes your manuscript on its own. If you keep Edit Mode suggestions or accept Proofing Pass corrections, those are the machine-suggested changes in the text; check your publisher's or contest's policy on machine-suggested edits.

Your drafts are never used to train models unless you opt in (see Privacy). Text reaches an AI service only when you trigger a feature, and only the passage or context that feature needs. We do not train on your manuscripts or personal notes unless you opt in, and we do not sell them.

4 · Disclosure and human oversight

We never hide that Robin is AI, and we never deny the technology if you ask. Robin is identified as an assistant at first interaction, and we do not write copy that implies a human produced something a model produced.

Human oversight is built into how text changes: the author writes the prose, and a change Robin suggests in Edit Mode, like a Proofing Pass correction, reaches the manuscript only when you keep it.

5 · Limitations we want you to know about

Reader Pulse scores are predictions, not verdicts. They are useful for finding where a draft loses a reader and are not a substitute for human beta readers or a developmental editor — they change when you spend those people, not whether you need them. Voice analysis describes patterns in your prose; it does not judge quality. Treat all of it as an informed second opinion.