At last year's RSAC Conference, Agentic AI dominated the conversation. This year, the conversation has changed.
The focus now isn’t just about what’s possible, it’s about how to safely put it into practice. As Agentic AI agents move into production environments, the challenge is no longer whether they can operate inside the enterprise—it’s how to govern what they can access and what they’re allowed to do.
These agents introduce a new class of identities that request access dynamically, inherit permissions through integrations, interact with other agents outside your control, and operate at machine speed.
Most identity programs weren’t designed for identities that behave this way.
Because of this new class of identities, this year’s theme at RSAC Conference 2026, Power of Community is especially timely. The industry is still working out how AI identities should be secured, and those answers are taking shape through shared experience, across practitioners, partners, and the broader security community.
RSAC 2026 will bring these conversations to the forefront, exploring practical approaches to governance, visibility, and privilege management for AI agents.
AI agents are forcing identity programs to adapt
AI agents don’t behave like traditional identities. They operate across systems, act autonomously, and evolve over time. That makes it harder to answer fundamental governance questions, including what an agent can access, why it has that access, and whether that access is still appropriate.
In real environments, this often shows up as access that expands gradually and indirectly. An agent designed for one task may gain additional permissions through integrations or reused workflows. Over time, access that was once appropriate may no longer reflect the agent’s intended role, and without clear visibility, those changes can be difficult to detect.
These challenges are no longer theoretical. They’re already shaping how organizations think about identity governance, and they’re a big part of the conversation heading into RSAC this year.
Exploring practical approaches to AI identity security
At RSAC, Delinea will lead sessions that reflect the real challenges organizations face in securing agentic AI.
Securing Intelligent Agents: How Delinea Enables Safe Adoption of Enterprise AI
Tuesday, March 24 | 10:20 – 10:50 AM PDT
This session explores how to anchor AI activity to identity and privilege controls. Matt Lawson, managing director and global identity leader from Protiviti, will join me, and together we will unpack how consistent identity governance across humans, machines, and AI agents reduces risk without slowing innovation. This session offers practical frameworks for balancing AI adoption and security in real environments. Register or learn more.
Agentic AI on Trial: Human Identity or Machine Identity?
Wednesday, March 25 | 9:40 – 10:30 AM PDT
This session takes the debate further. Chris Hickman, Chief Security Officer from Keyfactor, will join me to examine whether AI agents should be governed like human users, like machines, or whether a new model is needed. In this interactive session, we’ll ask the audience to help shape the discussion, providing a rare opportunity to see the community’s thinking in real time. Reserve your seat.
Identity security and infrastructure access
As AI agents take on more operational responsibilities, they’re increasingly interacting directly with infrastructure, systems and data. Access to servers, databases, and services is no longer limited to human administrators.
Delinea recently acquired StrongDM; a critical step in securing identities. Combining StrongDM’s just-in-time runtime authorization capabilities with Delinea’s identity security platform enables organizations to govern who—or what—can access critical systems in real time. For AI agents, this ensures that infrastructure access is dynamically controlled, auditable, and limited to what’s necessary, reducing the risk of persistent privileges or unintended exposure.
Identity security anchored to infrastructure access provides a unified control plane for both human and non-human identities, enabling enterprises to scale AI adoption safely while maintaining operational assurance.
Join the conversation at RSAC
RSAC has always been where the security community comes together to share ideas, challenge assumptions, and shape practical solutions. As AI continues to evolve, those conversations are more important than ever.
Every major shift in technology forces the industry to adapt. AI is no exception. The approaches organizations take today to secure intelligent agents will shape how safely these technologies can be adopted in the years ahead. That’s the real power of community: shared challenges leading to shared solutions. RSAC is where those solutions begin.
Visit us at booth N-4235 to explore Delinea’s approach to securing both human and non-human identities. View more information about our event presence here.
