Beyond the Model: The Expanded Attack Surface of AI Agents (Feb. 26th)
Thursday, February 26th, 2026: 1:00 PM to 2:00 PM
While AI red teaming has focused on stressing large language models through prompt injection and jailbreak attacks, the emergence of autonomous agents has fundamentally expanded the attack surface beyond what model-level security can address.
Virtual
This session moves beyond conventional LLM vulnerability assessment to examine the systemic security challenges introduced when AI systems gain agency, tool access, and the ability to interact with external systems and other agents.
Hosted by Dark Reading