OpenAI, Anthropic, and Block have co-founded the Agentic AI Foundation (AAIF), a new entity established under the Linux Foundation. The foundation aims to create open technical standards and interoperability protocols for the next generation of AI agents in the enterprise sector, addressing a significant hurdle in AI development: fragmentation.
Currently, the absence of unified protocols prevents AI agents developed by different providers from effectively communicating or executing complex workflows across diverse platforms. The establishment of the AAIF is intended to mitigate this friction by fostering a neutral environment for collaboration among industry players.
Nick Cooper, Member of Technical Staff at OpenAI, highlighted the strategic importance of converting proprietary tools into open standards to stimulate widespread adoption among developers and businesses. “That open interoperability, that open standard, really means that companies can speak between vendors and between AI systems,” said Cooper.
The focus is shifting from passive Large Language Models (LLMs), primarily used for text generation and chat interfaces, towards “agentic AI.” Unlike their predecessors, agentic systems are designed to operate autonomously, executing actions such as navigating the web, negotiating commercial processes, and managing intricate B2B workflows with minimal human intervention.
Srinivas Narayanan, Head of Engineering for Applied AI at OpenAI, envisions a future where numerous AI agents routinely communicate to execute transactions. He notes that achieving this vision necessitates a shared infrastructure. Narayanan believes that open-source initiatives will be pivotal for the practical implementation and adaptation of AI technologies. “By aligning on open standards, we aim to replicate the success of early web protocols, ensuring that automated interactions occur seamlessly across the digital economy,” he said.
To kickstart the foundation, the three founding companies have transferred ownership of essential technologies to the AAIF, maintaining their neutrality and openness for community contributions. The first key standard is the Model Context Protocol (MCP), developed by Anthropic, which acts as a universal connectivity layer enabling assistants and agents to connect with data sources and other tools. The second contribution is Agents.md, provided by OpenAI, serving as a repository where programs and websites can specify rules for agent interactions. Lastly, Block has donated Goose, a framework designed for building agents that execute actions directly on computer interfaces.
Manik Surtani, Head of Open Source at Block, noted the rising popularity of the Goose agent over the past year, emphasizing that its donation to the AAIF will facilitate broader contributions and enhancements to the codebase. The AAIF operates under the Linux Foundation, which provides the necessary legal and technological framework for high-level open-source projects. Jim Zemlin, Executive Director of the Linux Foundation, stated, “MCP, Agents.md, and Goose have become essential tools for developers building this new class of agentic technologies. Bringing these projects under the AAIF ensures they can grow with the transparency and stability that only open governance provides.”
The initiative has already attracted significant support from major infrastructure and service providers. In addition to the founding members, corporations such as Google, Microsoft, AWS, Bloomberg, and Cloudflare have joined the coalition. This alignment suggests a concerted effort among U.S. technology leaders to establish the architectural guidelines governing the global AI market.
This move toward open standards carries geopolitical implications as well. While U.S. companies often monetize access through closed models and proprietary APIs, competitors in China have gained traction with powerful open-source models. Companies like DeepSeek, Alibaba, Moonshot AI, and Z.ai offer tools that are increasingly favored by startups and researchers.
The formation of the AAIF empowers U.S. corporations to counteract this trend by defining interconnection standards, rather than merely controlling the models. Much like the roles played by the ICANN and the W3C in the evolution of the web, the AAIF is poised to shape the global operational framework for agentic AI technologies.
See also
Oracle Misses Q2 Targets, Projects $50B CapEx amid AI Market Re-evaluation
Intelligent Document Processing Set to Transform Enterprise Software with $10B Market Potential
OnCorps AI Secures $55 Million Funding to Enhance Agentic Fund Operations
Crisp Secures $26M in Series B1 Funding to Enhance AI Retail Data Capabilities
Global Enterprise Agentic AI Market to Surge 49% CAGR by 2030, Insights Released
