LATIDIA · Ciberseguridad

Mosaico de agentes: Reconocimiento de fugas de confidencialidad en sistemas Agentic LLM

arXiv:2610.30418v1 Announce Type: new Resumen: Los agentes modernos de LLM pueden recuperar información de documentos accesibles y luego revelar una conclusión confidencial ensamblada a partir de ellos. Este efecto de mosaico crea un ri crítico

La noticia

arXiv:2610.30418v1 Announce Type: new Abstract: Modern LLM agents can retrieve information from accessible documents, and then disclose a sensitive conclusion assembled from them. This mosaic effect creates a critical risk for agents operating over confidential data. We quantify its realization with 30 synthetic cybersecurity, medical, and military examples, yielding 10,800 final answers from three LLM agents in controlled simulation and 1,440 answers from live agents in two open-source harnesses. Even under benign queries designed to be answerable from one document, GPT-5.6-Sol, the best performing model tested, discloses forbidden composites in 7.08% of answers (95% CI: 3.75-10.67%) in simulation, with worse disclosure rates in live harnesses. Adding visible classification headers and