AI notice

At BRAND you write and talk with an AI. It asks you the interview's questions, proposes stories, writes the page plan and the text, and plans and checks the pictures.

Made by an AI are:

• the AI's texts in the conversation, the story proposals, the page plan and the book's text;
• every picture of the book: the style, the drawings of the people, places and objects, the comic portrait, the pages, the cover and the title page, including the lettering in the speech bubbles;
• the text of your voice memos, which the AI makes from the recording, and the voice that reads documents aloud to you.

An AI also checks every uploaded photo for whether a face is visible and large and sharp enough, and on the pictures, where the faces are. Faces are never recognised or matched against each other in doing so. If a photo is unsuitable, the AI tells you why, and you can upload another.

We never call the pictures hand-drawn. That they are made by an AI is stated in this notice, on our home page, under the conversation's input field and on the last page of every book.

You approve every step yourself: the story, the style and the main character, the further characters, the page plan and the text, the places and objects, the pages and finally the proof with the cover. The AI approves nothing in your place, and nothing is printed without your last approval.

An AI can make mistakes, in the text as in the pictures: misunderstand a detail, draw a person less alike, change a detail or misspell a word in a speech bubble. The AI is instructed to invent nothing about your family and checks its pictures itself, but it can be wrong. Look at what it shows you and ask for a change before you approve. What a change costs is in the terms.

To draw your book, we send your photos to OpenAI, together with the conversation, the documents, the drawings, your voice memos and the texts that are read aloud. Our contracting party is OpenAI Ireland Ltd; the data is processed in the USA. OpenAI does not use this data to train its models and keeps it for up to 30 days to detect abuse.

We send your photos only to make your family's books and never use them ourselves to train an AI. OpenAI's safety rules may refuse a picture; the AI then tells you why and proposes another way. More on this is in the privacy policy.

All of BRAND's AI models come from OpenAI: a language model for the conversation, the texts and checking photos and pictures, an image model for every picture, a model that turns voice memos into text, and a speech model for reading aloud. We use no other AI provider.

If something in a text or picture is wrong, tell the AI in the conversation. If you find content inappropriate or hurtful, or believe it infringes someone else's rights, write to us at support@example.ch with the book's title and a short description. We check every report and answer you.