Inside the Movement Defining AI Security: AI Agent Security Summit Now On-Demand

I’m still buzzing from the AI Agent Security Summit in San Francisco just a few weeks ago! From hallway discussions between sessions to rooftop chats overlooking the Bay Bridge, the excitement was contagious. Everyone came ready to learn, share ideas, and challenge what we thought we knew about securing AI. One thing became very clear to us all - we are no longer asking what to secure in AI. We’re defining how to secure it.

A Turning Point for AI Security

Sessions at the summit covered a wide range of topics that showed experts are already running experiments, developing policies, examining frameworks and so much more. They shared lessons learned from red team exercises, discussed how to treat autonomous, adaptive technologies, dug into the evolution of promptware, and talked through exercises in probing orchestration layers, memory recall, and tool integrations. And that was just from a handful of sessions. You could feel the energy across both tracks at the Commonwealth Club.

These conversations marked a turning point and a mindset realignment. The focus has shifted from theory to practice, from “why” to “how”. The summit made it clear - AI Agent Security isn’t a subtopic, it IS AI Security.

A Community Defining the Category

The best part of the summit? The incredible speakers, sessions, and insights - and the engagement from the hundreds of attendees. As Michael Barguy, Zenity CTO and co-founder said, “The most valuable thing we built wasn’t a model or a tool - it was a community.” He’s right. The real power of this summit is in the connections, the conversations, and the shared commitment to shape this discipline together. Now you can be part of that community and experience the insights, learnings, and research that made the event so impactful!.

Make it a “Purple Zentini” Moment

We’re thrilled to make all 17 sessions available to you on-demand so you can dive into the discussions shaping the future of AI security that made the event so valuable. Featuring voices from Google, Microsoft, Walmart, Stanford University, OWASP, ServiceNOW, OpenAI, and more - each session dives deeper into the real challenges and solutions defining AI Security (now available to watch at your own pace).

To relive the energy (or recreate the fun), try bringing the summit home with our signature drink:

Zentini Cocktail: Empress 1908 Indigo Gin, lemon juice, honey syrup, lemon twist

Zentini Mocktail: Pomegranate juice, lemon juice, honey syrup, sparkling water

Now grab your popcorn, pour a Zentini, and hit play - diving into the future of AI Security from the comfort of your chair (can’t wait to hear what you think)!

Looking Ahead

As Amanda Grady, VP of AI Platform Security at ServiceNow said it best: “It’s not one and done. Security is not set it and forget it – it absolutely has to be continuous.”

That mindset captures exactly what makes this community so special - it’s not about a single event or milestone. It’s about continuously learning, experimenting, and shaping what AI Security means in practice.

The conversations we started in New York and continued in San Francisco are only the beginning. I’m excited for what’s next. The future of AI Security is here, and it’s agent-centric.