Live Webinar: New Research on How AI Agents Act, Access, and Operate at Runtime

Join us September 29 | 11 AM PT / 2 PM ET to learn how to build an authorization model for AI agents that scales. 

Join Paul Webber, Principal Cybersecurity Industry Analyst at Software Analyst Cyber Research (SACR) and former Gartner Senior Director, Analyst, and Matt Wilson from Britive for a live practitioner session on what identity and access need to become as AI agents move into production. 

AI agents are calling tools, requesting credentials, accessing cloud and SaaS systems, and taking actions on behalf of people. As their numbers grow, the approval models identity teams have relied on start to break down. You cannot put a human in front of every action when agents operate continuously and at machine speed. 

But removing the human isn’t the answer either.

Authenticating an agent once doesn’t mean every action that follows should be trusted. The harder question is: should this agent be allowed to take this action, with this access, in this context, right now? 

Less AI theory. More practical answers. 

Paul will share what SACR learned developing its new ARISE research, Agentic Runtime Identity Security Enforcement (ARISE), and why agentic identity security needs to move beyond provisioning and one-time approval toward decisions and enforcement while agents are actually working. 

Then we’ll put it into practice.  

Through real scenarios and live demos, you’ll see: 

• What happens when an agent requests privileged access 
• Where agents can act autonomously and where humans still belong 
• How access can be scoped to a task and disappear when the work is done 
• What happens when an agent steps outside what was intended 

You’ll leave with a practical model for deciding what agents can do autonomously, where human approval still belongs, how to eliminate standing privilege, and how to enforce decisions at runtime. 

Bring your hardest questions. This one is live.