Securing the Agentic Frontier : Architecture, Risk, and Practice

While people are fixating on the threat AI brings to cybersecurity as
a discipline, there’s the considerably more pedestrian but crucial
perspective of securing these systems. The role of the CISO is
expanding beyond security practices to include risk management,
compliance, and more and these role expansions will cascade through
the organization.

Agentic systems collapse traditional boundaries between user, system,
and adversary. Not only do they introduce new risks, they challenge
foundational assumptions in security engineering. Identity and
authorization grants are fluid based upon contextual composition,
behavior becomes non-deterministic, and threat models become
continuous, not static.

What are the risks, challenges, and opportunities that are unleashed
at the intersection of security postures, agents with agency, ambient
authority, context-specific interactions, and engineering discipline?
How does one grow this capability in an organization and how does one
align their career in this direction.