Linux Foundation Launches Agentic AI Foundation to Standardize Open Agent Ecosystems

The Linux Foundation has announced the formation of the Agentic AI Foundation (AAIF), a new initiative designed to ensure AI agents evolve as open and interoperable technologies rather than fragmented, proprietary products. The foundation will serve as a neutral hub for open-source frameworks and protocols that enable agents to securely connect to tools, data, and applications at scale.

The AAIF launches with contributions from leading AI developers and cloud providers. Anthropic is donating its Model Context Protocol (MCP) to support standardized tool integrations; Block is contributing Goose, its agent framework used internally for development work; and OpenAI is adding AGENTS.md, a lightweight specification for guiding AI behavior inside code repositories. Founding members also include AWS, Bloomberg, Cloudflare, and Google, reflecting a broad industry commitment to trusted agent infrastructure.

The foundation will focus on shared safety principles, cross-platform orchestration, and compatibility guidelines, enabling enterprises and developers to adopt autonomous systems without creating custom connectors for every environment. By coordinating open standards around agent design and deployment, the Linux Foundation aims to build a sustainable foundation for the next generation of AI-driven productivity and automation.