The Agentic Artificial Intelligence Foundation is emerging as a central effort to build open infrastructure for agentic Artificial Intelligence applications. Founded in December 2025 under the Linux Foundation, the Agentic Artificial Intelligence Foundation serves as a hub for developing open infrastructure for agentic Artificial Intelligence. The foundation is focused on tools and standards intended to support interoperability, including servers such as the Model Context Protocol and Agents.md, while also encouraging broader development and experimentation around agentic systems.
Jim Zemlin, CEO of the Linux Foundation, said the larger agentic stack is expected to take shape within the foundation’s ecosystem. He pointed to the need for more than protocols alone, saying organizations deploying agents require servers like MCP, as well as gateways, orchestration layers and controls to keep agents aligned. The goal is to create a practical open framework that businesses can use as agentic systems become more central to software development.
Zemlin also said enterprises seeking to build safe agents may gain confidence from tools developed within the foundation because they are being co-developed by major software companies. That collaborative model is meant to give businesses a stronger base for adopting agentic infrastructure while reducing concerns about fragmentation across emerging tools and platforms.
Mazin Gilbert, executive director of the Agentic Artificial Intelligence Foundation, said the developer role is changing significantly as these systems mature. He described a shift away from writing and debugging code directly and toward working with protocols and capabilities that allow agent systems to adapt and handle concerns such as horizontal or vertical scaling. In that model, development work becomes less about manual coding and more about managing, connecting and guiding increasingly autonomous software systems.