What this collects

Five things, all about the device’s own security settings:

• Whether full-disk encryption is turned on.
• Whether endpoint protection software is installed and running.
• Whether operating system security updates are up to date.
• Whether the screen locks automatically after a short idle period.
• Whether the device’s firewall is enabled.

That is the entire list. It does not grow without the people running this system reviewing and publishing a change here first.

No exceptions, not a setting anyone can turn on:

• Keystrokes, or anything you type.
• Screenshots, screen recordings, or webcam/microphone access.
• The contents of files, or their names.
• Browsing history, or what any website shows you.
• Email, messages, or the content of anything you send or receive.
• Your location, beyond the fact that a company-managed device exists.

The agent answers five yes-or-no questions about the device itself. It has no way to see what you do on it, because it is not built to look.

Whether each check passes or fails, and when the device last checked in — visible to your employer’s administrators in MiraShield, the same way a failed security check on a company laptop would be. Nobody at Miranium AI, the company that builds MiraShield, sees this as a matter of routine; support only reaches it if your employer asks for help troubleshooting a specific problem.

A device that stops checking in for an extended period is flagged the same way a failed check is — not because something is assumed to be wrong, but because a monitoring program that cannot tell the difference between “fine” and “not reporting” is not one worth having.

This page describes the software. Questions about why it is being used, or what happens with what it reports, are for your employer to answer — they are the ones who enrolled the device, not Miranium AI.