Why AI Agents Need an Identity Layer, Not Just Guardrails
-
Nov 3rd
-
1:00-2:30 PM EST
-
Virtual
Event Details
This virtual executive roundtable, co-hosted by Okta and Deloitte, will bring together a curated group of senior technology, security and AI leaders to discuss the critical need for a new approach to identity and security in the era of agentic AI.
Identity models were built for humans logging in and static service accounts sitting quietly in the background. Neither was designed for AI agents acting autonomously on someone's behalf, making decisions, accessing systems and taking actions without a person in the loop. As agentic AI moves from pilot to production, the gaps in how organisations govern, authorise and track non-human identities are becoming impossible to ignore. This is a candid, off-the-record conversation among peers, no presentations, no sales pitches, just an honest exchange of what is working and what is not.
Sponsor
Okta
The Auth0 Platform is a customer identity and access management solution that’s highly extensible, customizable, and built to be secure by default. With Auth0, your development teams can build, secure, and seamlessly orchestrate identity with less time and effort. Meet customers' expectations for highly available, convenient, and trustworthy experiences. Business outcomes facilitated - increase efficiencies, Auth0 reduces engineering overhead associated with building and maintaining customer identity. Strengthen cybersecurity - Auth0 offers security solutions and integrations for threat mitigation across all identity threat vectors before, during, and after login. Increase revenue - Auth0 lets you build the pixel perfect, personalized customer experiences your business needs, so you can acquire and retain customers while growing their lifetime value.
Agenda
What's The Agenda?
This exclusive virtual roundtable brings together senior security and IT leaders navigating the growing complexity of agent identity, non-human access governance and agentic AI risk. In a focused and confidential setting, you will connect with peers to share strategies, challenges and perspectives on what it actually takes to govern AI agents across their full lifecycle.
We will explore where traditional identity models are failing in agentic environments, what is holding back a proper fix, and which part of the agent lifecycle is proving hardest to get right in practice.
Key discussion topics will include:
Where the Model Breaks Down
Traditional identity models were built for humans and static service accounts, not for agents acting autonomously. Where have you seen that model break down, and what did it expose?
The Problem with Inherited Credentials
Agents often inherit a human's credentials or share a generic service account. What breaks when an agent does not have an identity of its own, and what is holding back the fix?
Guardrails vs Restrictions
Blocking agent access is the easy call, but it usually just pushes activity into the shadows. Where is the line between a guardrail people work with and a restriction they work around?
The Hardest Part of the Agent Lifecycle
Managing an agent across discovery, authorisation and governance is a different challenge at each stage. Which of the three is hardest to get right, and why?
This roundtable is a chance to connect with like-minded leaders, compare notes on what is working, and leave with practical perspectives to bring back to your team, all from the convenience of a focused 90-minute virtual session.