Executives from Linux Foundation, Anthropic, Block, and OpenAI announcing the Agentic AI Foundation at a San Francisco press conference.
Executives from Linux Foundation, Anthropic, Block, and OpenAI announcing the Agentic AI Foundation at a San Francisco press conference.
Image générée par IA

Linux Foundation forms Agentic AI Foundation for open AI agents

Image générée par IA

The Linux Foundation announced the creation of the Agentic AI Foundation (AAIF) on December 9, 2025, in San Francisco, to foster open-source development of AI agents. Co-founded by Anthropic, Block, and OpenAI, the initiative includes donations of key projects: Anthropic's Model Context Protocol (MCP), Block's goose framework, and OpenAI's AGENTS.md. The foundation aims to promote interoperability and prevent fragmentation in AI agent technologies.

The Agentic AI Foundation (AAIF), launched under the Linux Foundation's umbrella, seeks to provide a neutral platform for advancing autonomous AI systems that can make decisions and coordinate across tools. This move addresses the shift from conversational AI to agentic technologies, ensuring transparent and collaborative evolution through open governance.

Anthropic contributed its Model Context Protocol (MCP), released in November 2024, which has become a standard for connecting AI models to external tools, data, and applications. Over 10,000 MCP servers are now active, with adoption by platforms including Claude, Cursor, Microsoft Copilot, Gemini, VS Code, and ChatGPT. 'MCP started as an internal project to solve a problem our own teams were facing,' said Mike Krieger, Chief Product Officer at Anthropic. 'A year later, it's become the industry standard... Donating MCP to the Linux Foundation as part of the AAIF ensures it stays open, neutral, and community-driven.'

Block donated goose, an open-source, local-first AI agent framework released in early 2025, which integrates language models with extensible tools using MCP standards for reliable workflows. 'We're at a critical moment for AI,' noted Manik Surtani, Head of Open Source at Block. 'By establishing the AAIF... we're building the infrastructure for an AI future that benefits everyone.'

OpenAI provided AGENTS.md, released in August 2025, a markdown-based standard for guiding AI coding agents in repositories, adopted by over 60,000 open-source projects and tools like GitHub Copilot and Cursor. 'For AI agents to reach their full potential, developers and enterprises need trustworthy infrastructure,' stated Nick Cooper, Member of the Technical Staff at OpenAI.

Platinum members include AWS, Anthropic, Block, Bloomberg, Cloudflare, Google, Microsoft, and OpenAI, with additional gold and silver tiers featuring companies like Cisco, IBM, and Salesforce. The AAIF will host events like the MCP Dev Summit in New York City on April 2-3, 2026. Jim Zemlin, Executive Director of the Linux Foundation, emphasized: 'Bringing these projects together under the AAIF ensures they can grow with the transparency and stability that only open governance provides.' This collaborative effort aims to standardize protocols, enhance safety, and support widespread adoption of agentic AI.

Ce que les gens disent

Reactions on X to the Linux Foundation's launch of the Agentic AI Foundation are predominantly positive, celebrating the open-source donations of MCP by Anthropic, goose by Block, and AGENTS.md by OpenAI to promote interoperability in agentic AI. Major companies including Google, Microsoft, AWS, Cloudflare, Snowflake, Docker, and others announced membership, emphasizing neutral governance and community-driven standards. Developers and analysts expressed excitement over preventing fragmentation and standardizing AI agents, with minimal skepticism noted.

Articles connexes

Microsoft's Build 2026 keynote presentation on AI agents and models in San Francisco.
Image générée par IA

Microsoft annonce des agents et des modèles d'IA à la conférence Build 2026

Rapporté par l'IA Image générée par IA

Microsoft a ouvert sa conférence de développeurs Build mardi en mettant l'accent sur les outils et agents d'IA. La présentation à San Francisco a mis en avant plusieurs nouveaux modèles et concepts matériels.

La Linux Foundation a annoncé son intention de créer la Tokenomics Foundation, qui élaborera des normes ouvertes pour mesurer les coûts de consommation de jetons (tokens) d'IA. Cette initiative est menée en partenariat avec la FinOps Foundation et bénéficie du soutien de douze organisations majeures. Elle sera officiellement lancée la semaine prochaine à San Diego.

Rapporté par l'IA

Plusieurs grandes entreprises technologiques ont rejoint une nouvelle fondation visant à développer des normes communes pour les systèmes d'intelligence artificielle.

Ubuntu et Fedora s'apprêtent à introduire des fonctionnalités d'intelligence artificielle dans un avenir proche. Cette décision survient alors que Fedora fait face à des réactions négatives de la part de certains utilisateurs concernant ses projets en matière d'IA. Les responsables ont affirmé qu'il n'y avait aucun signe de départ des utilisateurs en raison de ces changements.

Rapporté par l'IA

Anthropic aurait accepté de payer 200 milliards de dollars à Google au cours des cinq prochaines années pour obtenir l'accès à des puces et à des serveurs cloud. Cet accord, rapporté par The Information, fait suite à une entente précédente accordant au créateur de l'IA Claude l'accès à l'infrastructure de Google. Il met en lumière les investissements massifs qui alimentent le secteur de l'IA.

Ce site utilise des cookies

Nous utilisons des cookies pour l'analyse afin d'améliorer notre site. Lisez notre politique de confidentialité pour plus d'informations.
Refuser