The Linux Foundation has launched the Agentic AI Foundation (AAIF), a collaborative hub for developing agentic AI, featuring contributions from Anthropic, OpenAI, and Block. Major tech companies, including Google and Microsoft, support the initiative, which focuses on creating open standards for AI agents. Central to this effort is Anthropic’s Model Context Protocol (MCP), which has quickly become the standard for connecting AI models with external tools and applications, boasting over 10,000 servers. Block contributes “goose,” an open-source framework for AI agents, while OpenAI adds AGENTS.md, a widely adopted format for directing AI coding agents. The AAIF aims to prevent fragmentation in AI systems by promoting shared standards, ensuring interoperability between different frameworks and tools. As AI agents advance, these standards are essential for fostering collaboration, enhancing functionality, and avoiding isolated systems akin to the early internet.
Source link
