Anthropic has donated its Model Context Protocol to a new initiative under the Linux Foundation, as major technology companies move to formalise open standards for so-called agentic artificial intelligence.

The protocol, known as MCP, will become a founding project of the Agentic AI Foundation, a directed fund under the open-source umbrella group. The new foundation has been co-founded by Anthropic, Block and OpenAI, with support from Google, Microsoft, Amazon Web Services, Cloudflare and Bloomberg.

MCP is an open standard that links AI applications with external tools, data sources and systems. Anthropic introduced it a year ago and has positioned it as a shared interface for AI “agents” that interact with software and online services.

Wide industry adoption

Anthropic said use of MCP has expanded across AI products, infrastructure providers and corporate deployments during its first year.

There are now more than 10,000 active public MCP servers in use. These range from developer tooling to implementations inside large corporate environments.

Major AI platforms have integrated the standard. MCP has been adopted by ChatGPT, Cursor, Gemini, Microsoft Copilot and Visual Studio Code, along with other AI-based products.

Cloud providers have also added support. Deployment options for MCP now exist on AWS, Cloudflare, Google Cloud and Microsoft Azure.

Anthropic has expanded its own use of the protocol. Its Claude assistant now has a directory of more than 75 connectors that run over MCP.

The company has also updated its API. It recently introduced Tool Search, and added Programmematic Tool Calling for production-scale MCP deployments. These features manage large numbers of tools and reduce latency in complex agent workflows.

Growing technical ecosystem

MCP now has a community-run registry that lists available servers using the standard. Developers can browse and discover connectors and services that support the protocol.

A recent specification release added new functions. The November update introduced asynchronous operations and statelessness, and included features for server identity and official protocol extensions.

Software development kits for MCP now exist in all major programming languages. Anthropic reported more than 97 million monthly SDK downloads across Python and TypeScript.

The company has repeatedly described MCP as open-source, community-driven and vendor-neutral. It said those principles guided the decision to transfer the project into an external foundation.

Linux Foundation role

The Linux Foundation will steward MCP through the Agentic AI Foundation structure. The non-profit oversees many large open-source projects, including the Linux kernel, Kubernetes, Node.js and PyTorch.

It focuses on long-term project sustainability, shared technical infrastructure and neutral governance. Its model relies on open collaboration by companies and individual contributors.

The Agentic AI Foundation will sit within this framework. It is set up as a directed fund inside the Linux Foundation with backing from several large technology firms and financial services group Bloomberg.

The new body will support open standards and shared tooling for agentic AI systems. Its remit includes strategic investment, community development and common specifications for how AI agents operate.

Under the new structure, MCP joins two other founding projects. Block is contributing its “goose” project, and OpenAI is contributing its AGENTS.md specification.

Anthropic said grouping these projects together in the Agentic AI Foundation would promote shared work on standards and core tooling for AI agents. It said this would also support long-term neutrality and community oversight.

Governance and future work

MCP’s governance arrangements will stay in place despite the transfer. The current maintainers will continue to run the project. They will keep existing processes for community input and transparent decision-making.

Anthropic framed the move as part of a broader effort around open-source infrastructure for AI agents. It said open software and shared standards are important for security, innovation and public-interest uses of the technology.

“Open-source software is essential for building a secure and innovative ecosystem for agentic AI. Today’s donation to the Linux Foundation demonstrates our commitment to ensuring MCP remains a neutral, open standard. We’re excited to continue contributing to MCP and other agentic AI projects through the AAIF,” said Anthropic.

The Agentic AI Foundation will now oversee further development of MCP, goose and AGENTS.md as it brings additional projects into its portfolio.