Everyone is Deploying AI Agents. Almost Nobody Knows What They're Doing

AI agents are operating inside your enterprise; querying databases, triggering workflows, and taking action through APIs. As AI agents are adopted, organizations cannot see, track, or control what these agents are actually doing.

In this session, Roey Eliyahu, Co-Founder and CEO of Salt Security, challenges the industry’s narrow focus on LLM safety and exposes the much larger, invisible attack surface created by agentic systems.

Today’s security tools focus on the “brain.” But AI agents don’t just reason, they act. They operate across three interconnected layers: LLMs (reasoning), MCP servers (execution), APIs (action).

Roey introduces the Agentic Security Graph, a new framework for understanding how agents actually operate across your environment and where security breaks down.

Learn more at salt.security