But while Jack Dorsey’s Block has deep roots in the open-source community from its association with Bitcoin, OpenAI and Anthropic aren’t exactly champions of transparency. Anthropic’s flagship large language model (LLM) Claude, is closed-source, although the firm has an open-source project called Model Context Protocol (MCP) that allows LLMs from different companies to communicate.
Read more: This Former Cash App Executive Is Sounding the Alarm on Bitcoin
Similarly, OpenAI, despite its name, keeps ChatGPT closed-source. The firm infamously morphed into a for-profit company after starting out as a pure non-profit in 2015. It pivoted to a “capped-profit” status in 2019 before finally establishing OpenAI Global in 2023, a fully for-profit holding company.
“When Anthropic and OpenAI are in the same foundation announcement, pay attention,” one X user said in response to Dorsey’s post about the announcement.
Although the coalition is somewhat unusual, all three companies are forging ahead with the new initiative. Block will contribute a new agentic AI framework dubbed “goose” to the foundation. Anthropic will offer its popular MCP protocol, and OpenAI will provide a project called “Agents.MD,” which is a tool used by coding agents.
“Agentic AI will be open, accessible, and community-driven,” Block’s announcement states. “The world has repeatedly proven that open collaboration produces better outcomes than competition in isolation.”
FAQ ⚡
-
What is the Agentic AI Foundation?
It’s a new initiative by Block, OpenAI, and Anthropic to support open-source projects focused on autonomous AI systems.
-
Why are these companies partnering?
They aim to accelerate safe, interoperable development of agentic AI by pooling tools, frameworks, and research.
-
What technologies will each group contribute?
Block is offering its new “goose” framework, Anthropic is providing MCP, and OpenAI is adding its Agents.MD project.
-
How will this impact real-world AI applications?
The collaboration is expected to advance autonomous systems used in areas like self-driving, fraud detection, and customer support.