OpenAI, Anthropic, and Block just launched the Agentic AI Foundation under the Linux Foundation, transferring ownership of key AI agent technologies to create open standards. The move signals a major shift toward interoperable AI systems that can communicate across platforms, potentially reshaping how businesses deploy autonomous AI agents.
The AI industry just made its biggest bet yet on autonomous agents working together. OpenAI, Anthropic, and Block announced they're cofounding the Agentic AI Foundation under the Linux Foundation, transferring ownership of critical AI agent technologies to create industry-wide standards.
The foundation immediately gains control of three essential tools that developers already use to build AI agents. Anthropic's Model Context Protocol (MCP) lets agents connect and interact with each other, while OpenAI's Agents.md allows websites to specify rules for coding agents. Block is contributing Goose, its framework that taps multiple language models to perform computer tasks.
"MCP is used by many companies, but there are others [who don't use it]," Nick Cooper, who leads protocol work at OpenAI, told Wired. Making these tools open standards should encourage broader adoption and integration across the industry.
The timing reflects a crucial shift happening right now. While chatbots dominated 2023, the focus is moving toward AI agents that actually do things - book flights, manage transactions, negotiate deals. These systems need to talk to each other seamlessly, and that's where standardization becomes critical.
Srinivas Narayanan, OpenAI's CTO of B2B applications, envisions armies of AI agents routinely communicating in business settings. "Open source is going to play a very big role in how AI is shaped and adopted in the real world," he said in the Wired interview.
But there's more at stake than technical interoperability. The foundation already has backing from Google, Microsoft, AWS, Bloomberg, and Cloudflare - essentially creating a US-led consortium to define how AI agents operate globally.
This comes as Chinese AI companies like DeepSeek and Alibaba are releasing powerful open source models that developers increasingly prefer. Some experts worry this could give Chinese firms a strategic advantage, making the new foundation's work even more significant for US competitiveness.
Block's head of open source Manik Surtani says the company's Goose agent has "soared in popularity" over the past year. By donating it to the foundation, anyone can contribute to the codebase and build new capabilities on top.
The Linux Foundation's Jim Zemlin called the transferred technologies "essential tools for developers building this new class of agentic technologies." The foundation will provide the neutral governance these tools need to scale globally.
What makes this particularly clever is how it positions the founding companies. While the standards themselves are technically neutral, if these tools become the global default for AI agents, OpenAI, Anthropic, and Block gain enormous influence over how AI gets deployed worldwide.
The parallels to web standards are striking. Just as ICANN and W3C shaped internet evolution, the Agentic AI Foundation could determine how autonomous AI systems interact across the global economy. The difference is that this time, US companies are getting ahead of the standardization process instead of playing catch-up.
The Agentic AI Foundation represents more than technical cooperation - it's a strategic move to ensure US companies shape the future of autonomous AI systems. As businesses increasingly deploy AI agents for everything from customer service to complex transactions, having standardized protocols becomes crucial. The real winners here might be the three founding companies, who get to influence global AI agent standards while appearing to champion openness. It's a smart play that could determine whether US or Chinese approaches dominate the next phase of AI deployment.
