JournzyES

Trust Center

Your manuscript is yours. Full stop.

Writing with a supporting AI can feel daunting, especially when nobody else has read what you are writing yet. This page tells you plainly what we do with your text, what we never do, and where you can check it for yourself.

Last updated: July 21, 2026 · Journzy is in private beta

The four promises our whole project rests on

These are not slogans. Each one is written into the code before it appears on this page, and below we show you how to check it.

We do not sell your data

Never. Not your text, your email, or your activity. There are no advertisers, advertising cookies, or marketing third parties. The email you give us for the waitlist is used only to manage your access and launch communications.

We do not train AI on your work unless you turn it on

The permission starts off and stays that way until you decide otherwise. Your work is stored as your working document, but while permission remains off, no manuscript line or story-bible card is kept in the training record: there is no second copy intended to train models.

The AI reads to help; it never writes your story

Journzy watches the continuity and consistency of your manuscript and warns you when something does not fit. Every flag includes an “Intentional” button. Nothing changes on its own: you decide and make every edit.

Your text never appears in our analytics

We measure clicks and performance, not writing. A record may say that you opened a screen or how long a feature took; it never contains the text that was inside. The editor, story-bible cards, and AI flags are protected so their content never travels in those records. It is forbidden by design.

Training: off by default and in your hands

This is the question that matters most to an author, so we answer it honestly. We do not train models on your ideas or projects unless you want us to and tell us explicitly. You have to enable the permission for your resolved checks to help improve Journzy's AI.

  1. Click your name at the very bottom of the left column, in the Account & settings area.
  2. Open the Data & privacy tab.
  3. There you will find the Help improve the AI with my accepted suggestions switch.

If you ever turn it on, the permission remains narrow: only AI checks you have already resolved — the flag and your decision — are used, never your drafts beyond what the AI reviewed to warn you about a discovery or inconsistency. You can turn it off whenever you like, without explanation; from that moment, your work is excluded from any future use.

Why you can hold us to this

The permission is off by default in the database itself. While it remains off, the training record only says that the content was withheld: the columns where writing could appear stay empty. This is not a policy we promise to follow; it is how the system is built.

How your text moves through Journzy

Being honest means telling you what does leave your screen, not only what does not. When the AI checks your story's continuity and consistency, this is what actually happens:

  1. It stays in Europe. Your manuscript and story bible are stored on Journzy-managed infrastructure, on Hetzner servers in Germany and Finland. Images are stored in Cloudflare R2 under European jurisdiction; Cloudflare also protects delivery of the service.
  2. Only what is needed is sent for each check. Journzy gathers the passage being reviewed and the relevant part of your story bible, then gives them to the model running the check. The whole work is not sent when that context is enough to assess the flag.
  3. The engine runs with zero data retention. Today models are served through OpenRouter with Zero Data Retention enforced for the whole account. Text may remain ephemerally in memory while the response is calculated — like a page held open while it is read — but it is not kept as history or used for training.

What never accompanies your text

Your identity does not travel with the writing: the engine receives the context it needs to check, never your name or email. Our analytics and error logs do not receive your manuscript either.

To learn whether Journzy works, we store things such as a screen name, an internal identifier, a count, or a duration. PostHog processes that metadata in its United States region under the EU–U.S. Data Privacy Framework and standard contractual clauses; it does not receive the content of your work.

Exporting, deleting, and exercising your rights

We will be clear, including about what still needs polish: during the private beta, preparing a copy of your data and deleting your account are requests we handle manually.

Write to hola@journzy.com and a person on the team will prepare a copy of the data you have entrusted to us or process the deletion of your account. Automatic controls do not exist yet; when they are ready, we will update this page.

The GDPR gives you the rights of access, rectification, erasure, portability, restriction, and objection. We respond within the legal one-month deadline. If you believe we have mishandled your data, you can complain to the Spanish Data Protection Agency.

Deletion cannot be undone

We manually review the systems linked to your account, erase your manuscript, story bible, and images, and confirm when the process is complete. If any information must be kept for a period because of a legal obligation, we will tell you explicitly. Before requesting deletion, ask for a copy of any data you want to keep.

The fine print, which we do not hide

Everything you have read on this page also appears in our formal documents, written in the same plain language we use with you:

Is there a question this page does not answer? Write to hola@journzy.com; a person on the team will reply.