Search papers, labs, and topics across Lattice.
This paper formalizes the concept of reconnaissance-driven pentesting for AI agents, emphasizing the need to uncover hidden weaknesses through a systematic knowledge extraction process. The authors introduce the Know Your Agent (KYA) framework, which automates the probing of AI agents to build comprehensive target profiles that inform more effective indirect prompt injection attacks. Evaluated against agent-security benchmarks and a real-world coding agent, KYA demonstrates significant improvements in attack efficacy, underscoring the importance of reconnaissance in AI security assessments.
AI agents are vulnerable to indirect prompt injection attacks, and KYA reveals how targeted reconnaissance can significantly enhance pentesting effectiveness.
Traditional pentesting uses reconnaissance at each step to uncover unseen weaknesses, build stronger attacks, and advance the objective; we argue that AI agents require the same treatment. We formalize agent reconnaissance by modeling the process and identifying the knowledge assets it seeks to extract: what they are, how they are used, and which agent weaknesses they exploit to give adversaries leverage in indirect prompt injection attacks. We instantiate these insights in Know Your Agent (KYA), a framework that automates black-box, reconnaissance-driven pentesting by probing agents, building target profiles, and using those profiles to craft stronger attacks. We evaluate KYA on agent-security benchmarks and a real-world coding agent, and release KYA, its benchmarks, and baseline implementations for reproducibility.
