Skip to content
VirtualOnlinePast event

AI TechDay: Agent Smith

Event artwork for AI TechDay: Agent Smith

This event has ended

This listing is kept as an archive. Explore upcoming events to find your next one.

See upcoming events
About this event

What to expect

πŸ€– Agent Smith: From Exploitation to Containment β€” a deep dive into autonomous AI agents that don't just generate text, they act. This two-part online session from Global AI Karlsruhe takes you from offensive IoT exploitation to enterprise-grade defensive controls, all powered by autonomous AI agents making real-time decisions on real hardware.

Part 1: Agent Smith Gets Hardware β€” Autonomous IoT Hacking From Debug Port to Cloud API

Speaker: Marc Plogas

Watch as autonomous agents attack a real off-the-shelf IoT device with zero human intervention. Using MCP-driven tool orchestration running on a Raspberry Pi, the agents autonomously probe hardware debug interfaces, extract firmware secrets, and intercept cloud traffic through a rogue WiFi access point. No manual hacking, no memorized commands, no Kali cheat sheets β€” just AI agents making their own decisions about what to probe, what to extract, and what to flag.

This isn't theoretical. Every demo runs on open-source tools, and you'll see what the agents find, what they miss, and what happens when a device fights back. Whether you build IoT products, secure enterprise networks, or just want to know what your smart devices are doing behind your back, this session will make you uncomfortable.

Part 2: Agent Smith Gets Governed β€” Securing AI Agents with Microsoft Foundry, Control Planes & Guardrails

Speaker: Dr. Alexander Wachtel

After breaking things, the harder challenge: how do we control autonomous agents? Modern AI agents call APIs, orchestrate tools, access data, and make decisions β€” making them powerful and dangerous. This session explores how to secure them using Microsoft Foundry, the Foundry Control Plane, and built-in guardrails, treating agents as active actors in your infrastructure.

Through real scenarios, you'll learn how to:

  • Enforce identity and access boundaries
  • Restrict tool usage and execution paths
  • Prevent data exfiltration
  • Monitor and intervene in real time

You'll also see where guardrails fail β€” because autonomous systems don't always break in predictable ways.

What to Expect

  • Live demos with real hardware and real agents
  • Autonomous attacks and defensive controls
  • Practical insights into AI agent security
  • Honest results: what works, what doesn't

Who Should Attend

This session is essential for security professionals and red/blue teams, architects and IT leaders, developers and AI engineers, and IoT/embedded engineers who need to understand

Running an AI event?

List it here, free.

Meetups, workshops and conferences go in front of thousands of builders after a quick review.

List your event