> For the complete documentation index, see [llms.txt](https://help.steeple.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://help.steeple.com/en/assistant-ia/assistant-ia-confidentialite.md).

# Privacy, security and permissions

This article describes what the AI Assistant can see, what it can do, and how your data is processed.

## User rights compliance

The assistant acts **always within the rights scope of the signed-in person**. It cannot:

* Access content that the user does not have access to in Steeple
* Perform an action that the user could not perform themselves
* Bypass the roles, permissions, or moderation rules in force

## Confirmation of sensitive actions

Any action that **creates, modifies, or deletes** a piece of data triggers a **explicit confirmation request** :

* A card details the planned action, whose fields can be edited before validation
* Two buttons allow accepting or declining
* No write action is performed without validation

Read-only actions (search, reading) do not require confirmation.

## Data sent to the model

To reply to your message, the assistant sends the AI model:

* The content of your message
* The files you attached (extracted text, image)
* The information strictly necessary for the request (for example, the name of a searched-for colleague)
* The context of the current conversation

No data is used to train models.

## Conversation storage

* Conversations are **stored** and viewable from the page **My conversations** ; starting a new conversation does not erase the previous one
* Each conversation is only visible to the person who conducted it
* A conversation can be **deleted** individually from the list
* Attached files are not **stored** beyond message processing

## Usage limits

Daily usage is capped per user, depending on the organization's plan. This cap protects the infrastructure; it does not imply any additional retention of your data.

## Improvements and quality

Feedback **thumbs up / thumbs down** that you leave on the assistant's responses are used for evaluation and quality improvement purposes. They may be accompanied by an optional comment.