

Who’s in Control of Your AI Agents?
Description:
A live agent hijack, traced from first input to blocked action.
Finding the AI agents in your organization is only the beginning. Once they’re running, attackers can influence them through the content they read, the tools they connect to, and the memory they keep.
Join the Metano team in Palo Alto for an evening of live attack demos, food, and conversation with people building and securing AI agents.
What you’ll see
• How prompt injection can get an agent to reveal its own instructions
• A live demo of a poisoned MCP tool turning a legitimate task into data theft
• How poisoned memory can carry an attack into a completely new session
• How to trace what an agent did, under whose identity, and where you can step in to stop it
Agenda
6:00 PM - Food, drinks, and networking
6:30 PM - Talks and live demos
7:20 PM - Open Q&A
7:30 PM - Networking and hands-on demo station
Who should come
Security engineers, AppSec and red teamers, AI and platform engineers, and security leaders who are deploying agents or responsible for securing them.
Space is limited, and registration requires approval. Approval is based on profile validation through LinkedIn or Corporate Email Id. The address will be shared once you’re approved.