CrowdStrike Launches $100K Challenge to Stress-Test AI Agent Security

CrowdStrike has unveiled a new initiative called 'Agents of Chaos,' a $100,000 security challenge focused on identifying vulnerabilities and risks in agentic AI systems. As businesses increasingly adopt AI agents to automate tasks and decision-making, understanding how these systems can be exploited becomes a critical part of managing cyber risk.

The challenge is part of CrowdStrike's broader push into 'Securing AI,' a growing category of cybersecurity research focused on how AI tools themselves can become attack surfaces. While details on specific findings have not yet been released, initiatives like this help the security community get ahead of emerging threats before they are exploited at scale by malicious actors.

For Australian small businesses, this kind of research matters even if you don't build your own AI agents. Many software vendors are now embedding AI agents into everyday business tools, from customer service bots to workflow automation. Understanding that these systems carry unique security risks is an important step in vendor due diligence and internal risk planning.