What I Found When I Started Breaking AI Agents
LLM agents have the same vulnerability classes as web applications, plus entirely new ones. Here's what I'm finding in the field.
ai-securityllmoffensive-securityowasp
Writing on API security, offensive security automation, and security engineering.
LLM agents have the same vulnerability classes as web applications, plus entirely new ones. Here's what I'm finding in the field.
Organizations have thousands of API endpoints, most undocumented. Here's why traditional pentesting misses them and what to do about it.