AI RESEARCH
NAAMSE: Framework for Evolutionary Security Evaluation of Agents
arXiv CS.AI
•
ArXi:2602.07391v2 Announce Type: replace AI agents are increasingly deployed in production, yet their security evaluations remain bottlenecked by manual red-teaming or static benchmarks that fail to model adaptive, multi-turn adversaries. We propose NAAMSE, an evolutionary framework that reframes agent security evaluation as a feedback-driven optimization problem. Our system employs a single autonomous agent that orchestrates a lifecycle of genetic prompt mutation, hierarchical corpus exploration, and asymmetric behavioral scoring.