LATIDIA · Ciberseguridad

Black-Box Red-Teaming de sistemas multiagente a través del aprendizaje por refuerzo

arXiv: 2610.28966v1Tipo de anuncio: nuevo Resumen: Los agentes del modelo de lenguaje grande (LLM) se implementan cada vez más en sistemas multiagente (MAS) para realizar tareas complejas. Los ataques previos de equipos rojos de caja negra se centran principalmente en si

La noticia

arXiv:2610.28966v1 Announce Type: new Abstract: Large language model (LLM) agents are increasingly deployed in multi-agent systems (MAS) to accomplish complex tasks. Prior black-box red-teaming attacks mainly focus on single agents, but we find that these methods are far less effective in MAS, where multiple sub-agents may not directly interact with the user. Therefore, we introduce ReMAS, the first reinforcement learning-based red-teaming framework tailored to MASs, which fine-tunes attacker LLMs to generate effective adversarial prompts for system prompt extraction. The framework follows a two-step process: first, a rewrite stage refines base attack prompts to increase extraction success, and second, a template generation stage constructs attack templates that improve the likelihood of