Ethical and Responsible AI
In Transept the AI suggests and a person decides. Every document records which was which, and nothing you write is used to train a model.
No cardFirst 5 pages freeYour work never trains AI
Literess, the AI editor in Transept
You: Make Grandma grumpier.
Literess: Done. She doesn’t just say it anymore. She grumbles it. sagte → raunzte
Ask in plain words. You approve every change. Meet Literess →
Literess, our AI editor, proposes each change and waits for your click. Proofreading fixes can be reviewed one by one, with Accept and Undo, and a workflow can stop at a review step until someone signs off. If you turn on Allow All, her changes apply without the click, and they are still listed as hers.
We don't use your documents to train models, and neither do the AI providers we use. That holds on every plan, the Free plan included.
Every document keeps an activity record: what a model wrote, and what a person changed, accepted or approved. The originality report counts it by word and turns it into a PDF.
Our team doesn't read customer documents in routine work. For a support case, we ask you to send us the document you need help with. Our logs never contain your document text.
We don’t sell ways to hide AI use or promise text that passes as human-written. When a publisher, a journal or a client asks whether AI was used, the answer is yes, and the originality report shows what you changed and reviewed.
Transept is run from Prague, and your documents are stored in Germany. Every company that handles your data is listed on our sub-processors page, with what it does and where.
A translation run writes the first version, steered by your glossary and style guide.
Her edits and fixes arrive as suggestions for you to accept or decline.
Each change has Accept and Undo, and every version, the AI's and yours, is kept.
Before every export, Literess asks whether you read the translation through.
Who wrote what, by word, with your dated confirmation, in the originality report.
The people you write for are asking how AI was used. Every Transept document keeps the answer: what the AI drafted, what a person changed, and what a person checked. The originality report puts it on one page, so the work you did is visible next to the AI's.
Two reports on the Free plan, unlimited on every paid plan.
Originality report
Chapter 3 · Ukrainian → English
Reviewed in full by the author, confirmed on Oct 21, 2026
| Question | Answer |
|---|---|
| Is it used to train AI? | No. Not by us, and not by the AI providers we use. |
| What does an AI model see? | Only the passage being worked on, for the length of that request. The providers (Anthropic, OpenAI, Google and others) are named on the security page, and none of them trains on it. |
| Who at Transept reads it? | No one, in routine work. For a support case, we ask you to send us the document you need help with. |
| What do your logs contain? | Technical events: the model, timing, cost and errors. The text of your documents is never written to them. |
| Where is it stored, and for how long? | In Germany, on Hetzner's ISO 27001 data centre in Falkenstein, until you delete it. Deleting your account deletes your content. |
| Who owns the translation? | You do. Our terms claim no ownership of your text or its translation. |
EU AI Act. From August 2026, text that AI generated or changed and that is published to inform the public on matters of public interest has to be disclosed, unless a person reviewed it and someone holds editorial responsibility for it. Transept's review steps and the originality report help you document that review.
Amazon KDP. KDP counts a book translated with an AI tool as AI-generated, even after substantial edits, and asks you to declare it. Declare it. The report shows what you changed and reviewed, and it doesn’t replace the declaration. Our AI-detector test shows why a detector score is no guide either way.
Journals, universities and clients. Many accept AI help on condition that a person checks the result, and ask you to say how it was used. The report is a record you can attach.
This is general information, not legal advice. Whether a rule applies to your text is for you, and your publisher or client, to judge.
Your text is never used to train AI, not by us and not by any provider we use, and it stays out of our logs. Delete your account and your text is deleted with it.
How we handle dataHosted in Germany, in Hetzner’s ISO 27001-certified data center. Every sub-processor is listed, and a DPA is one email away.
All sub-processorsGlossaries, styleguides, translation memory and per-string notes go with every request. You set them up once per project.
GlossariesThe AI drafts; you decide. Review every change side by side, keep or revert it, and send a review link when a client needs to look.
Review linksThe full API on every paid plan. Connect Claude or another AI assistant over MCP, or start runs from your own tools.
Developer guideLarge catalogs, custom workflows, a DPA for procurement. Transept stays self-serve; write to us for the rest.
support@transept.aiFeatured on
