Sanctuary Browser Guard adds a protection layer to ChatGPT and Gemini web sessions.
Before a prompt is submitted, the extension sends the draft text to a user-configured Sanctuary deployment for preview, redaction, or blocking. This helps teams prevent raw sensitive data from leaving the browser while keeping AI usage aligned with company security and privacy controls.
The extension is designed for organizations that want to use public AI interfaces without exposing raw customer, employee, health, financial, or other regulated data in prompts.
What it does: - intercepts prompt submission on supported pages - previews whether the prompt should be allowed, redacted, or blocked - shows a protected version when redaction is required - stores only local extension settings in browser storage - does not download or execute remote code
