Today, the Linux Foundation launched the Agentic AI Foundation with three founding projects: Anthropic’s Model Context Protocol (MCP), Block’s goose agent framework, and OpenAI’s AGENTS.md standard.
The foundation brings together the companies building the infrastructure layer for agents: Anthropic, Block, OpenAI, Amazon, Google, Microsoft, Cloudflare, and Bloomberg, alongside key tooling and platform companies.
Docker is joining as a Gold member.
From Open Source to Production
The timing reflects how quickly the space has matured. A year ago, MCP launched as an open source project from Anthropic, solving a specific problem: how AI systems connect to tools and data. It’s now running on 10,000+ public servers and adopted across Claude, ChatGPT, Cursor, Copilot, VS Code, and Gemini.
Six months ago, companies started deploying agents that take real actions, triggering builds, accessing databases, modifying infrastructure, executing workflows. That shift from prototype to production created new questions around protocols and governance.
Today, foundational protocols that helped answer those questions, protocols like MCP, are moving to the Linux Foundation under the same governance structure that stewards Linux and PyTorch.
Why Neutral Governance Matters
When infrastructure becomes critical, developers won’t build on protocols that could change arbitrarily. And larger teams and enterprises want shared standards.
Over the past year we’ve partnered with Anthropic, Block, and other key players in the AI ecosystem to help create and embrace standards like MCP, Goose, and AGENTS.md. The Agentic AI Foundation creates a structure for the industry to unite behind these standards, building an ecosystem of interoperable tools that benefit developers.
Docker is excited to join as an active Gold member to drive innovation in developer-first, secure tools across our ecosystem.
What Happens Next
The protocols exist. Adoption is happening. The foundation ensures these protocols evolve transparently, with input from everyone building on them.
Docker helped build that structure for applications. Now we’re doing it for agents.
Learn more at aaif.io
Quelle: https://blog.docker.com/feed/
