Guard: AI & Media Filter 作者： Elhio

Browse the web with peace of mind. Guard proactively flags AI-generated, violent, or explicit content on the pages you visit – keeping your browsing experience safe, private, and distraction-free.

WHAT IT DOES

• Automatic, on-device scanning: As you browse, Guard inspects images using content-credential metadata (C2PA), EXIF/XMP signals, and an optional lightweight AI model. No image data is ever uploaded for this automatic scan.
• Custom reaction: You are in control. Choose how flagged images are handled: mark with a badge, blur, or hide them entirely.
• Granular control: Easily toggle detection categories on or off and whitelist sites where Guard shouldn't run.
• Optional deep-scan: For uncertain results, sign in and click 'Verify' on a specific image to send just that image to our servers for a thorough second opinion. This process is never automatic.

PRIVACY

• Automatic detection is 100% on-device. Your browsing history and page URLs are never transmitted.
• Images only leave your browser/device when you manually initiate a 'Verify' request.
• We do not use third-party analytics, advertising, or tracking.
• Guard is open source (AGPL-3.0).

Ready for a safer browsing experience? Click ‘Install’ to get started – no account or sign-up required to begin blocking.