cloud-engineering
devops
agentic-ai
linux-foundation
mcp
open-source

Agentic AI Foundation

The AI Industry's Most Important Open Standard

On December 9, something remarkable happened: OpenAI, Anthropic, and Block—competitors in the AI race—donated their core agent technologies to the Linux Foundation. The Agentic AI Foundation is born.

What is the Agentic AI Foundation?

A vendor-neutral home for the protocols that will define how AI agents operate in enterprise environments.

On December 9, 2025, OpenAI, Anthropic, and Block—direct competitors in the AI race—took the unprecedented step of donating their core agent technologies to the Linux Foundation.

The Donated Projects

Model Context Protocol (MCP)

Donated by Anthropic

The "USB-C for AI"—a universal standard that connects LLMs to external tools, data sources, and APIs.

10,000+

Published MCP Servers

Universal

Tool Integration

Adopted

OpenAI, Microsoft, Google

AGENTS.md

Donated by OpenAI

A markdown standard for defining AI coding agent behavior—how they should interact with repositories, what permissions they have, and how they should operate.

60,000+

Open Source Projects

Markdown

Simple, Universal Format

Adopted

Cursor, Copilot, Devin, Gemini

goose

Donated by Block

An open-source, local-first AI agent framework with MCP-based integration for trusted agentic workflows.

Local-First

Privacy by Design

MCP Native

Built-in Integration

Open Source

Fully Auditable

Founding Members

The heavyweights of AI and cloud infrastructure, united under one foundation:

AWS

Cloud

Anthropic

AI

Block

Fintech

Bloomberg

Finance

Cloudflare

Infrastructure

Google

AI/Cloud

Microsoft

AI/Cloud

OpenAI

AI

Why This Matters for Infrastructure Teams

1. Security Through Transparency

Open governance means you can audit the protocols AI agents use to interact with your systems. No black boxes—every interaction is defined by public, reviewable specifications.

2. Interoperability

Standard protocols let AI agents from different vendors work together without custom integration. Build once, connect to any compliant agent.

3. No Vendor Lock-In

The foundation ensures no single company dominates how agents connect to your infrastructure. Your integrations work across the ecosystem.

The Bigger Picture

"We are seeing AI enter a new phase, as conversational systems shift to autonomous agents that can work together."

— Jim Zemlin, Executive Director, Linux Foundation

What's Changing

  • From chat interfaces to autonomous agents
  • From single-vendor to multi-agent systems
  • From proprietary protocols to open standards

What Teams Should Plan For

  • AI agents as first-class infrastructure citizens
  • Standardized, auditable agent protocols
  • Multi-vendor agent orchestration

What This Means in Practice

For DevOps Teams

MCP servers become infrastructure components. You'll deploy, monitor, and secure them like any other service—but now with standardized protocols and cross-vendor compatibility.

For Platform Teams

AGENTS.md files define how AI interacts with your repositories. Platform teams can set organization-wide policies for agent behavior, permissions, and boundaries.

For Security Teams

Open protocols mean auditable agent interactions. Security teams can review, test, and validate how AI agents interact with systems—no more trusting vendor black boxes.

The Path to Open Agent Standards

Nov 2024

Anthropic launches MCP

Model Context Protocol introduces universal tool integration for LLMs

Early 2025

OpenAI adopts MCP

Major validation as competitor embraces the standard

Mid 2025

AGENTS.md gains traction

60,000+ projects adopt the markdown agent standard

Dec 9, 2025

AAIF Founded

Linux Foundation announces Agentic AI Foundation with founding members

Competitors United for Open Standards

When OpenAI, Anthropic, Google, and Microsoft agree on something, infrastructure teams should pay attention. The Agentic AI Foundation signals that AI agents are moving from experimental to enterprise-ready.

For DevOps and platform teams, this means planning for a world where AI agents are first-class citizens in your infrastructure—with standardized, auditable protocols.

Is your infrastructure ready for agentic AI?

Learn more at talk-nerdy-to-me.com/news