OpenAI, Anthropic and Block Form Foundation to Advance Action‑Based AI Agents
Why does this matter? OpenAI, Anthropic and Block have announced a joint foundation dedicated to advancing AI agents that act rather than just converse. The collaboration, first reported under the headline “OpenAI, Anthropic, and Block Are Teaming Up to Make AI Agents Play Nice,” brings together two of the most prominent AI research labs with a venture firm that has backed several generative‑AI startups.
While the technology promises tools that can book flights, manage calendars or even negotiate deals on a user’s behalf, the move also surfaces regulatory questions about accountability, cross‑platform interaction and the use of public web resources. The foundation’s focus on programs that take concrete actions marks a departure from the chat‑centric models that have dominated recent releases. Yet, standards for how these autonomous agents should behave remain unsettled, and lawmakers are only beginning to grapple with the implications.
This development signals a shift toward a potentially lucrative new paradigm in which AI agents use the web and negotiate with one another to power all sorts of applications.
The new foundation reflects a nascent shift from chat-based AI systems to greater use of programs that take actions on behalf of users. This kind of agentic AI promises a potentially lucrative new paradigm in which AI agents use the web and negotiate with one another to power all sorts of applications. Consumers may, for example, use AI assistants to buy and book things, while businesses use AI agents to manage transactions and customer interactions.
Srinivas Narayanan, chief technology officer of B2B applications at OpenAI, envisions a time when large numbers of AI agents routinely communicate with one another in the course of business. The AI industry working across the same open standards should help ensure that those interactions happen seamlessly. "Open source is going to play a very big role in how AI is shaped and adopted in the real world," Narayanan says.
Will these standards stick? The Agentic AI Foundation now houses Anthropic’s Model Context Protocol, OpenAI’s Agents.md, and the Goose framework, all under a shared open‑source umbrella. By pooling these tools, the three founders hope to smooth interactions between autonomous agents and the web.
Yet the shift from chat‑based models to action‑oriented programs is still in its early days, and it is unclear whether developers will adopt the proposed norms at scale. Because the foundation is newly formed, its governance and enforcement mechanisms remain opaque. The promise of agents that negotiate with each other to power applications sounds attractive, but practical deployment hurdles have not been detailed.
Moreover, the financial incentives hinted at in the quote are speculative until real‑world use cases emerge. In short, the collaboration marks a concrete step toward coordinated agentic development, though the path to widespread, interoperable AI agents is far from guaranteed. Stakeholders will likely watch how the foundation’s open‑source releases integrate with existing ecosystems, and whether they can foster reproducible benchmarks for agent behavior.
A test of interoperability.
Further Reading
- Linux Foundation announces Agentic AI Foundation joined by Anthropic, OpenAI and Block - SiliconANGLE
- Linux Foundation Announces the Formation of the Agentic AI Foundation (AAIF), Anchored by New Project Contributions Including Model Context Protocol (MCP), Goose, and AGENTS.md - PR Newswire
- OpenAI co-founds the Agentic AI Foundation under the Linux Foundation - OpenAI
- Linux Foundation forms Agentic AI Foundation to be new home for MCP, goose, and AGENTS.md - SD Times
- Newly launched Agentic AI Foundation to take charge of MCP’s future - Techzine Europe
More in Policy & Regulation
Trump signs executive order threatening to punish states that pass AI laws
President Donald Trump signed an executive order this week that threatens to withhold federal funds...
Congress Raises Alarm as Section 702 Expands Beyond Original Terror Surveillance
Seventeen years after Section 702 was written into law, the very framework that was meant to target...
NY signs two bills requiring ads to disclose AI‑generated avatars
New York’s governor put his signature on two pieces of legislation Thursday, targeting a niche...
AI satellites need 300‑500 GW, requiring millions of units, new cooling tech
Why does the idea of AI‑powered satellites matter now? The push to run massive machine‑learning...
Deepseek uses thousands of smuggled Nvidia chips for AI training in China
Deepseek’s latest AI models are being trained on thousands of Nvidia GPUs that, according to...
Common Questions Answered
What is the primary goal of the foundation announced by OpenAI, Anthropic, and Block?
The foundation aims to accelerate the development of action‑based AI agents that can perform tasks like booking flights or managing calendars, moving beyond purely conversational chat models. By pooling resources, the three organizations hope to create a shared ecosystem that enables agents to interact with the web and each other more effectively.
Which open‑source tools are included under the Agentic AI Foundation?
The foundation houses Anthropic’s Model Context Protocol, OpenAI’s Agents.md specification, and the Goose framework. These components are placed under a common open‑source umbrella to standardize how autonomous agents communicate and execute actions on behalf of users.
How are consumers and businesses expected to use the new action‑oriented AI agents?
Consumers could rely on agents to automatically purchase tickets, reserve accommodations, or manage personal schedules, while businesses might deploy agents to handle transactions, process orders, and streamline customer‑service interactions. This dual use case illustrates the potential for agents to act as both personal assistants and enterprise workflow tools.
What concerns does the article raise about the adoption of the foundation’s standards?
The article notes that the shift from chat‑based models to action‑oriented programs is still in its infancy, and it remains uncertain whether developers will embrace the proposed norms at scale. Adoption will depend on how quickly the ecosystem matures and whether the standards prove practical for real‑world applications.