The Agentic AI Foundation has announced 43 new members, bringing its total membership to 190 organisations and adding new participants across infrastructure, payments, security, cloud, robotics, academia and the public sector.
The new cohort includes F5, GoDaddy, Stripe and TRON as Gold Members, alongside a wider group of Silver and Associate Members including Atlassian, Fastly, Teradata, VeriSign, Consumer Reports, the NSW Government, Pacific Northwest National Laboratory, Sandia National Laboratories, the University of Washington and the U.S. Army.
For enterprise buyers, the significance is not only the membership number. It is the type of organisations now gathering around open agent standards. As AI agents move from experiments into business workflows, the market is beginning to treat interoperability, identity, security, governance and infrastructure as shared problems that cannot be solved by model capability alone.
What was announced
The Agentic AI Foundation, a Linux Foundation project, describes itself as a neutral home for the open standard agentic AI stack. Its founding projects include Model Context Protocol, goose and AGENTS.md, with a focus on protocols and tooling that help agents work across platforms and systems.
On 18 May, the Foundation said it had added four Gold Members, 27 Silver Members and 12 Associate Members in the past quarter. The announcement frames the growth around production-grade agentic AI, with members spanning application delivery, payments, cybersecurity, cloud native development, robotics, enterprise technology, public sector organisations and universities.
That spread matters. Agentic AI is no longer a narrow software feature. A production agent may need to identify itself, connect to tools, retrieve data, trigger workflows, route traffic, handle payments, operate across cloud environments and leave a defensible audit trail. Those requirements cut across many parts of the technology stack.
Why open standards are moving up the buying agenda
Enterprise AI agents only become commercially useful when they can act across real business systems. That creates a practical challenge for buyers: every additional tool, connector, model, workflow and permission boundary increases the risk of fragmentation.
If each supplier builds its own isolated agent framework, buyers face higher integration costs, weaker portability and a more complex governance burden. Open standards are one way to reduce that risk. They give enterprises a clearer route to connect agents with systems, evaluate supplier claims and avoid locking critical workflows into a single proprietary architecture too early.
The Agentic AI Foundation announcement points to several areas where standardisation is likely to matter most:
- Interoperability: how agents connect to tools, data sources, workflows and other agents.
- Identity: how an agent is recognised, verified, permissioned and associated with a real organisation or owner.
- Security: how agent actions, tool calls and model interactions are protected and monitored.
- Governance: how organisations set policies, approve actions, track behaviour and create audit evidence.
- Infrastructure: how agents run reliably across cloud, edge, sovereign and hybrid environments.
For buyers, these are procurement questions as much as technical questions. A demo can show what an agent does in a controlled environment. A standards conversation helps determine whether that agent can survive contact with enterprise architecture, compliance teams and operational risk management.
What it means for suppliers
For agent builders and enabling-technology suppliers, the direction of travel is clear. Enterprise customers will increasingly expect products to fit into a broader agent ecosystem rather than operate as standalone islands.
That means suppliers should be ready to explain which standards they support, how their agents connect to external tools, how identity and permissions are handled, how logs and audit trails are generated, and how customers can switch models, infrastructure or orchestration layers without rebuilding everything from scratch.
There is also a commercial upside. Open standards can make adoption easier by giving buyers more confidence that an agent product will connect to existing systems and future platforms. For smaller suppliers, standards can help them compete inside larger enterprise accounts because integration becomes less dependent on bespoke work.
The adoption signal
The bigger signal is that agentic AI is becoming an ecosystem category. Infrastructure companies, security firms, payment providers, universities, public agencies and enterprise software vendors are all trying to shape the rules before the market hardens.
That does not mean every standard will win, or that every enterprise agent deployment will be open source. It does mean buyers should pay attention to the underlying architecture. The question is not simply which agent performs best today. It is whether the surrounding environment will remain governable, portable and secure as agent usage scales.
For enterprises moving beyond pilots, open standards are becoming part of risk reduction. They help buyers ask better questions, compare suppliers more fairly and build agent strategies that do not depend on one closed stack.
The Agentic Expo angle
Agentic Expo is focused on market-ready AI agents. The growth of the Agentic AI Foundation underlines what market-ready increasingly means: not just autonomous capability, but credible integration, governance, identity, security and interoperability.
That is why the category needs a dedicated B2B meeting place. Buyers need to understand the practical stack behind agents. Suppliers need to show how their products fit into that stack. Standards bodies, infrastructure providers, security teams, agent builders and enterprise adopters all have to be part of the same conversation.