Installation

Prerequisites

Browser Guard is available starting from the Business tier.

You will need a Project API Key (format: pv_live_...) from your dashboard under Settings → Security → Browser Guard.

For deployments of more than 5 users, use Enterprise Deployment to pre-configure the extension via MDM — no manual key entry required.

Platform	Interception Method	Enforcement
ChatGPT	Fetch	Block, Warn, Justify, Redirect, Modify
Claude	Fetch	Block, Warn, Justify, Redirect, Modify
Microsoft Copilot	WebSocket	Block, Warn, Justify, Redirect, Modify
Perplexity	SSE	Block, Warn, Justify, Redirect, Modify
Grok	Fetch	Block, Warn, Justify, Redirect, Modify
DeepSeek	XHR	Block, Warn, Justify, Redirect, Modify
Gemini	XHR	Block only (Protobuf encoding prevents modification)
Meta AI	Fetch	Block, Warn, Justify, Redirect, Modify

500+ AI platform domains are monitored for usage tracking and categorization. No prompt interception — visibility only.

The extension intercepts prompts at the DOM level, before they leave the browser. No network proxy, no TLS break-and-inspect.

User types a prompt on a supported AI platform.
The extension captures the prompt before submission.
The prompt is sent to your gateway for policy evaluation.
Based on the policy decision, the extension allows, blocks, warns, or modifies the prompt.
A color-coded toast notification informs the user of the outcome.

Toast Color	Meaning
Green	Prompt allowed (or modified with redactions)
Red	Prompt blocked — not sent to the AI platform
Amber	Warning — user may proceed
Blue	Justification required
Purple	Redirect to approved alternative

Chrome/Edge: Right-click the extension icon → Remove from Chrome.
Firefox: Go to about:addons → find the extension → Remove.
Enterprise deployments: Remove the force-install policy from your MDM. The extension is automatically uninstalled on next policy sync.