Transparency
How Stoory uses artificial intelligence
Stoory is a publishing platform that uses artificial intelligence systems to create and develop stories, series, and linked materials. This page explains where and how AI is employed, what its limits are, and how to report content to us.
Which features use AI
- Text generation: titles, synopses, narrative structure, chapters of stories, and episodes of series.
- Images and covers: cover artwork is generated by image models; the typographic part and the editorial brand are composed by our tools.
- Translations: editions in other languages are produced with automatic translation and quality controls.
- Editorial author portraits: the profile pictures of Stoory editorial authors are generated with AI.
- Promotional videos: when present, these are edited from materials of the work and may include AI-generated elements.
- Automated moderation: texts submitted by users and generated outputs undergo automated content security checks.
AI-generated and AI-assisted: they are not the same thing
AI-generated indicates content created to a substantial extent by the artificial intelligence system.
AI-assisted indicates content of human or editorial origin that has been modified, translated, refined, or supported by AI.
We use these two categories distinctly in our internal provenance records and in the metadata of the files we generate.
Fictional editorial identities
Some titles in the catalogue are signed by Stoory editorial authors: these are fictional narrative identities, not real people. Their respective portraits are generated with artificial intelligence. We indicate this explicitly on the author pages.
We do not present these identities as real people and we do not promise the imitation of the style of real existing authors.
Editorial review
Every work goes through the Stoory editorial process: automated consistency and quality checks, content security verifications, and, when necessary, manual review before or after publication. Human review is not systematic on every single page.
Limits and possible errors
Artificial intelligence systems can produce errors: factual inaccuracies, narrative inconsistencies, repetitions, typos, incorrect visual details, or imperfect translations. We do not promise perfection or infallibility of the systems we use.
Stoory content consists of entertainment and narrative works: they do not constitute medical, legal, psychological, educational, or professional advice of any kind.
Use of data
We use the data you provide (for example, the idea entered in the creation path) to produce the requested work and for security checks. The processing of personal data is described in the Privacy Policy.
The artificial intelligence systems used by Stoory are not used for social scoring, emotion recognition, or biometric classification of people.
Machine-readable file marking
Where technically possible, we insert machine-readable metadata into the files we generate that declares the use of artificial intelligence: in PDFs (document metadata), in EPUBs (provenance metadata in the package), and in PNG covers (provenance text chunks).
This marking is a basic measure: it is not a cryptographic provenance signature (for example, C2PA / Content Credentials) and it does not include invisible watermarking. For promotional videos, machine-readable marking is currently unavailable, while the visible disclosure remains present.
When a file reaches us already with the provider's provenance metadata, we do not remove it: we add our own.
How to report content
If you find content that is problematic, inaccurate, or that you believe violates third-party rights, please write to us from the Contact page, indicating the title of the work and the reason for the report. We evaluate every report and can correct, unpublish, or remove the content.
This page is updated when the systems or the purposes of using AI on Stoory change.
