OpenAI, Anthropic, and Block team up to make AI agents play well

OpenAI, Anthropicand roadblock We’ve co-founded a new open source organization – the Agentic AI Foundation – to advance standards artificial intelligence Agents.

The three companies are also transferring ownership of some widely used proxy technologies to the organization. This includes anthropologists Model Context Protocol (MCP)Which allows agents to communicate and interact; OpenAI agents.mdwhich allows programs and websites to define rules for coding agents; and GooseA A framework for construction agents developed by Block. These technologies were already free to use, but through the new foundation it will be possible for others to contribute to their development.

“MCP is used by many companies, but there are others[that don’t use it],” says Nick Cooper, who leads work on the protocol at OpenAI. Cooper says making MCP an open standard should encourage developers and companies to adopt it and build systems that integrate agentic AI. “This open interoperability — this open standard — really means that companies can talk across service providers and across proxy systems.”

The Agentic AI Foundation is set up within the Linux Foundation, which oversees development of the widely used open source Linux operating system As well as other projects. The foundation provides legal and technological support for creating open source organizations. Other companies that have signed on to the AAIF, besides the three founding members, include Google, Microsoft, AWS, Bloomberg and Cloudflare.

The new foundation reflects an emerging shift from chat-based AI systems to the increased use of software that takes actions on users’ behalf. This type of agent AI promises a potentially profitable new model in which AI agents use the web and negotiate with each other to run all kinds of applications. For example, consumers can use AI assistants to buy and reserve things, while businesses use AI agents to manage transactions and customer interactions.

Srinivas Narayanan, CTO for B2B applications at OpenAI, envisions a time when large numbers of AI agents routinely communicate with each other in the context of work. An AI industry that operates according to the same open standards should help ensure that these interactions occur seamlessly. “Open source is going to play a very big role in how AI is shaped and adopted in the real world,” Narayanan says.

The issue of openness seems crucial for AI right now. US companies mostly make money by providing access to powerful closed models through application programming interfaces, or APIs. Meta previously released the weights for its best model, Llama, so anyone can download and run it, though the company has recently signaled a shift to a more closed approach. A number of Chinese AI companies, including DeepSeek, Alibaba, Moonshot AIand Z.aiprovides powerful open source models that have become popular with developers, startups, and AI researchers. some They worry that this image could give Chinese companies a major strategic advantage over time.