I’m excited to introduce the first secure agentic browser: Prisma Browser. AI is quickly moving from assisting… to acting. And that changes everything. What happens when an agent has your credentials? When it can take actions across apps on your behalf? This isn’t just an evolution of copilots. It’s a completely new risk model. With Prisma Browser, we built a secure agentic workspace where AI can act. With full visibility, governance, and real-time protection at the moment of action. 🛡️ Secure every AI-driven action 🔒 Protect sensitive data in real time ⚖️ Govern both human and agent activity We also support a bring-your-own-LLM approach, giving organizations flexibility without compromising security. AI won’t just help us work. It will work for us. And it needs to be secured accordingly. Link to blog in the comments. Elad Gavra Ohad Bobrov Asaf Ovadia Yonatan Gotlib Reuven Eliyahu Chris Gaba Monique Lance

Ofer Ben-Noon “New risk model” is exactly the right framing. Once agents can act across systems with credentials, the problem shifts from access control to continuous decision governance what actions are taken, in what context, and with what constraints over time. Static permissions won’t be enough. You need real-time policy, observability, and the ability to intervene mid-execution. Feels like security is moving from guarding systems to governing behavior.

Like
Reply

Huge milestone, Ofer. Securing the browser has always been critical, but with the rapid adoption of AI agents and LLMs, it’s now the ultimate frontier. Looking forward to seeing how this reshapes secure productivity without the typical friction.

Like
Reply

wowwwww we're definitely living in the future of cyber security 🫨

This is where things get real. Agentic AI needs a completely new security model.

Like
Reply

Game changing launch - a secure agentic browser is exactly what the AI era needs

Like
Reply

that's one big leap!

Like
Reply
See more comments

To view or add a comment, sign in

Explore content categories