Anthropic’s Model Control Protocol (MCP), Google‘s Agent2Agent, and LLMs.txt are protocols emerging for AI-enablement, focusing on making websites accessible for AI.
Microsoft’s NLWeb, linked to the first-generation web protocols, allows easy integration of AI-powered conversational interfaces, essentially transforming websites into AI apps.
NLWeb utilizes existing structured data formats like RSS, enabling websites to become AI-powered through a process that leverages modern AI technologies.
NLWeb simplifies the integration of AI capabilities on websites by processing structured data, enhancing it with external knowledge using LLMs, and creating a universal natural language interface for users and AI agents.
Different AI protocols such as Google’s Agent2Agent and LLMs.txt serve specific purposes like agent communication and improved access to web content, distinct from NLWeb’s focus.
NLWeb offers advantages in controlling how AI systems interpret website content and interact, providing better navigation and understanding.
Enterprises like Chicago Public Media, Eventbrite, and O’Reilly Media are already leveraging NLWeb to enhance their websites and AI capabilities.
NLWeb presents a valuable opportunity for enterprises to optimize internal AI capabilities for tasks like data retrieval and content understanding.
While NLWeb is in early stages, leading-edge companies can pilot it for internal use, but substantial adoption may still take a few years to mature.
Enterprise AI leaders should consider NLWeb as a crucial technology for AI-enabling their websites, paving the way for enhanced user experience and AI interaction.