I've been selected for the first cohort of Agentic AI Foundation ambassadors.
What is the Agentic AI Foundation, and why does MCP joining it matter?
AAIF is a Linux Foundation initiative focused on open-source infrastructure for AI agents. One of its founding projects is the Model Context Protocol (MCP), created by Anthropic and now contributed to the foundation. That move matters more than it looks. It takes the piece that connects agents to the outside world and puts it under open, shared governance instead of a single vendor.
What does the response to the ambassador program tell us?
The response says demand for open agent infrastructure is running far ahead of what anyone planned. The program planned for 10 ambassadors. It received more than 1,000 applications. It ended up welcoming 138 people across 41 countries. A program that expected ten people and drew a thousand is a signal in itself. A lot of practitioners see agent infrastructure as something worth building in the open.
Why does MCP matter for agent security?
MCP matters for security because it's becoming one of the main ways agents reach tools, data and real workflows, and that connection is where the security questions start. What can an agent actually access. What is it allowed to do. And afterward, what happened, and what evidence someone can review.
That last part is the work I've been doing at Oktsec: security, controls and evidence for AI-agent work. So this isn't a side interest. It sits on top of the same problem, at the layer where the standards are being written.
What I want to bring to the program is practical. The things I keep running into building AI agents and working in cybersecurity, shared in the open, where others can use them.
You can see the ambassadors here: aaif.io/ambassadors.
Grateful to have been selected, and looking forward to the work.