CISO Summit

As organizations accelerate adoption of agentic AI, they face a paradox: the tools driving innovation create huge security blind spots. Traditional IAM frameworks and security controls are not built to manage the unique challenges posed by AI agents, which operate autonomously, acquire system-level access and execute actions at machine speed. With 81% of teams deploying AI agents but only 14% securing formal approval, risks escalate rapidly, leaving businesses vulnerable to breaches, compliance failures and operational disruptions.

The AI security paradox is fundamentally an identity governance problem. AI agents bypass traditional provisioning, operate across complex delegation chains and lack the audit trails regulators will soon demand. Addressing this requires a paradigm shift. Platform engineering teams must treat AI agents as a distinct identity class, applying policy-as-code frameworks, dynamic risk-based authorization, and full observability to every agent action. This approach helps businesses mitigate risks without stifling innovation, ensuring AI adoption aligns with security and compliance mandates.

Join Atulpriya Sharma, Principal Developer Advocate at Improving, CNCF Ambassador to uncover actionable strategies for governing AI agents and securing your organization’s future.

Key Takeaways:
- Recognize why AI agents represent a new, ungoverned identity class beyond traditional IAM capabilities.
- Learn how AI agents silently acquire system-level access and the risks this poses
Explore policy-as-code frameworks to govern agent behavior without hindering innovation.
- Discover observability techniques to create robust audit trails for regulatory compliance.
- Gain a practical governance blueprint to align platform and security teams effectively.