Behavry

AI usage monitoring and DLP scanning — Behavry agent governance

Behavry gives organizations visibility and control over how their people use AI services in the browser. The extension monitors a fixed list of AI service domains (ChatGPT, Claude, Gemini, Copilot, Perplexity, and others) and applies your organization's data-protection policy before content is submitted. It does not monitor general web browsing. It activates only on the enumerated AI service domains and stays inactive everywhere else. What it does: - Detects sensitive data (credentials, secrets, personal information) in prompts before they are sent to an AI service, using pattern matching and an optional in-browser machine-learning classifier - Applies your policy to flag or block submissions that contain sensitive data - Reports governance events to your Behavry platform tenant so administrators have a record of AI usage across the organization Privacy and data handling: - Data-loss-prevention analysis runs locally in the browser. Prompt content is not sent to Behavry for scanning. - The optional machine-learning classifier runs entirely on the user's device. No prompt text leaves the browser for classification. - The extension reports governance events and findings metadata to your organization's Behavry tenant for administrator visibility and audit. Behavry is the operational control layer for AI agents and AI usage. This extension covers the human-in-browser surface. Access requires enrollment through your organization's Behavry deployment.