What happened
Google launched an open-source Model Context Protocol (MCP) server, directly connecting AI agents to Chrome Enterprise APIs, to enhance browser security management for IT and security teams. This server allows agents to automate tasks like configuration reviews, Data Loss Prevention (DLP) rule creation, and alert investigations by calling Chrome Enterprise APIs. Users interact in plain language via the Gemini CLI or other MCP-compatible clients, with agent-created rules automatically prefixed 🤖 in the Admin Console. The reference implementation is available on GitHub, requiring a Chrome Enterprise Premium subscription for full DLP features.
Why it matters
This initiative reduces manual effort for IT and security teams by automating routine browser security tasks, providing faster visibility and configuration. Platform engineers gain a mechanism to integrate AI agents into existing security workflows, streamlining policy deployment across managed and unmanaged devices. While agents offer suggestions, human review remains critical for policy changes, particularly for DLP rules requiring Chrome Enterprise Premium. This follows Tilde's launch of reversible AI agent sandboxes, underscoring the need for secure agent deployment and oversight.
