OpenAI, Anthropic and Block create new body to standardise Agentic AI

Industry leaders transfer key technologies to a neutral foundation as agentic systems move into real-world use

by e4m Staff
Published: Dec 11, 2025 11:12 AM | 1 min read

OpenAI, Anthropic and Block have formed the Agentic AI Foundation (AAIF), a new open-source organisation under the Linux Foundation, aimed at developing interoperable standards for agentic artificial intelligence. The move is supported by Google, Microsoft, AWS, Bloomberg and Cloudflare, signalling broad industry alignment around shared infrastructure for the next generation of AI systems.

As part of the initiative, the companies are donating several widely adopted technologies to the foundation. Anthropic’s Model Context Protocol, used for enabling agents to connect and interact, OpenAI’s AGENTS.md format for providing project-specific instructions to coding agents, and Block’s Goose framework for building agentic tools will now be governed in an open, community-led environment. These technologies were already free to use, but their transfer allows developers across the ecosystem to contribute to their evolution.

AGENTS.md, introduced in 2025, has already seen adoption across more than 60,000 open-source projects and frameworks. Its inclusion in AAIF is intended to ensure long-term, vendor-neutral stewardship as agentic systems shift from experimentation to production.

The foundation will provide a neutral platform to maintain and advance interoperability standards, with oversight from the Linux Foundation, which brings long-standing experience in managing major open-source projects. By consolidating core protocols and frameworks under one umbrella, AAIF aims to prevent fragmentation and support the scalable, safe deployment of agentic AI across industries.