The Secret Life of AI Agents: What They Do When No One is Watching

Date: 9/29/2026

Time: 11:00 a.m. PT / 2:00 p.m. ET

Duration: 45 minutes

AI agents are changing the meaning of privileged access. They do not simply log in. They act, invoke tools, request permissions, and execute workflows that cross agentic, non-human, and human identities, often with no person watching. 

Software Analyst Cyber Research has just published the first analyst framework built specifically to address the AI access problem. ARISE, Agentic Runtime Identity Security Enforcement, boils agent security to one question: should this agent be allowed to take this action, with this tool, this credential, and this delegated authority, right now?

In this session, SACR introduces the framework and explains why the market is moving from access decided in advance to access decided, enforced, and proven at runtime.

Discuss the Drivers Behind the Framework

SACR and Britive work through what actually breaks when agents operate at machine speed, drawing on recent real-world cases:

Malicious MCP servers. Agentic coding tool containment. Supply chain compromise. Browser agent prompt injection.

Discover the Questions Practitioners Should be Asking Right Now

How much privilege should an agent hold between tasks?

What bounds an agent acting on a person's behalf?

What happens after access is granted and the agent does something no one anticipated?

And how does any of it work when enterprise systems span cloud, SaaS, and on-premises?

Attendees will leave with a practical checklist for agentic access: inventory agents, remove standing privilege, enforce at runtime, capture intent and action, and escalate high risk workflows to a human. 

Speakers 🎤:
Paul Webber, Software Analyst Cyber Research (SACR)

Principal Cybersecurity Industry Analyst

Hosted by:

Matt Wilson, Britive
Principal Product Manager

Register to join the event and be sure to bring your questions for answers live from the team!