The next version of the web will be built for machines, not humans (www.economist.com)

A paradigm shift in web interaction is on the horizon as AI and machine learning technologies evolve towards the development of the "agentic web." Inspired by Sir Tim Berners-Lee's vision for an intelligent web from 1999, this new framework aims to allow AI agents—powered by large language models (LLMs)—to act autonomously rather than merely generating text. Recent advancements, including the Model Context Protocol (MCP) developed by Anthropic, enable these agents to interact directly with various online services through standardized communication, bypassing human-centric application programming interfaces (APIs). The Agentic AI Foundation, launched by industry leaders like Anthropic, OpenAI, Google, and Microsoft, is set to develop open-source standards for AI agent interaction, emphasizing the urgent need for cohesive protocols in a competitive landscape. As emerging tools like Microsoft's Natural Language Web (NLWeb) start bridging conventional web content with agent-friendly interfaces, the implications for advertising and online engagement could be transformative. Marketers may need to adapt to a model where targeting focuses on "agent attention," moving beyond direct consumer interaction. While the potential to increase efficiency is vast—potentially expanding web usage exponentially—the challenges of miscommunication and security risks associated with autonomous AI actions remain critical concerns developers must address.