In a major step toward standardising and streamlining the rapidly growing field of agentic AI, OpenAI, Anthropic and Block—backed by Google, Microsoft, Amazon Web Services (AWS), Bloomberg and Cloudflare—have joined forces to launch the Agentic AI Foundation (AAIF) under The Linux Foundation. The initiative represents one of the strongest industry-wide collaborations to date, aimed at developing open, interoperable frameworks that ensure agentic AI systems function reliably and safely across diverse environments. As part of the launch, OpenAI is contributing AGENTS.md, described as “a lightweight standard for project specific instructions” that has already gained significant traction across more than 60,000 open-source projects and widely used developer tools including Codex, Cursor, Devin, Gemini CLI, GitHub Copilot and VS Code.
The formation of AAIF comes at a pivotal moment, as agentic AI rapidly matures from experimental prototypes into real-world production systems. The foundation is intended to prevent fragmentation in this fast–evolving ecosystem, establishing common protocols that support consistent performance, safety and interoperability. Much of this momentum builds on OpenAI’s technical contributions over the past year—including the Agents SDK, Apps SDK, the Model Context Protocol, and its gpt oss models—which have shown that open standards can achieve meaningful adoption across varied developer communities.
Operating as a directed fund under The Linux Foundation, AAIF will provide a neutral governance structure for these emerging interoperability standards. The Linux Foundation already stewards some of the world’s most critical open technologies such as the Linux Kernel, Kubernetes, Node.js and PyTorch, making it a natural home for a collaborative effort of this scale. With founding participation from major technology and cloud providers—including OpenAI, Block, Anthropic, Google, Microsoft, Amazon, Bloomberg and Cloudflare—the foundation aims to guide agentic AI development in a direction that balances innovation with public interest.
By creating a shared space for defining standards, AAIF seeks to ensure that the next wave of agentic AI systems can integrate smoothly across platforms, reduce operational risks and enable developers to build more powerful and trustworthy AI-driven agents.
