Agents hunting Agents: Open Source + AI approaches.

​An agent swarm is making moves. Is it launching an attack, or simply doing its job? JoinAgents hunting Agents on October 19 at the Silicon Valley AI Hub in Menlo Park, where you’ll explore how to distinguish malicious AI behavior from legitimate agent activity.

​You’ll work in teams, with mentoring from leaders in AI and cybersecurity, to use AI to spot malicious AI behavior. Along the way, you’ll explore Vigil, a 100% open source way to use agents and other AI to tell the difference between malicious and benign agents. You’ll also consider how an agent harness can help keep your own agents behaving as intended.

​Some understanding of AI agents and cybersecurity will be helpful. Bring your questions and be ready to examine ambiguous behavior, compare interpretations with your team, and decide what evidence would make you trust an agent, or flag it for closer scrutiny.