Google Chrome ships WebMCP in early preview

Google Chrome has introduced an early preview of WebMCP (Web Model Context Protocol), a new web standard designed to transform how AI agents interact with websites. Developed collaboratively by engineers at Google and Microsoft and incubated through the W3C Web Machine Learning community group, WebMCP enables websites to expose structured, callable tools directly to AI agents via a browser API called navigator.modelContext. This approach allows AI agents to access and perform actions on websites more effectively by interacting with clearly defined tools rather than navigating traditional user interfaces built for humans.

WebMCP eliminates the need for developers to maintain separate back-end MCP servers for connecting web applications with AI platforms. Instead, they can wrap existing client-side JavaScript logic into agent-readable tools, which enables AI-driven automation within a user’s current browsing session without complex re-architecting. For example, a travel company could use WebMCP to allow AI agents to search, filter, and book flights directly on its site, improving speed, reliability, and precision.

By standardizing how websites communicate with AI agents, WebMCP supports smoother integration and broader AI usage in everyday web interactions. Its rollout in Chrome 146 marks a significant step toward enabling seamless AI-agent-enabled experiences on the web, opening new possibilities for innovation in areas like customer support, e-commerce, and task automation.