LATIDIA · Ciberseguridad

Un punto de referencia a gran escala y una evaluación de riesgos de los ataques de análisis de tráfico en los servicios de LLM en la nube

arXiv: 2609.31877v1Tipo de anuncio: nuevo Resumen: Los servicios de modelo de lenguaje grande (LLM) basados en la nube crean un canal lateral de tráfico a nivel de red que puede exponer el comportamiento del modelo, el indicador y la tarea a pesar del cifrado. Desde el paquete

La noticia

arXiv:2609.31877v1 Announce Type: new Abstract: Cloud-based Large language model (LLM) services create a network-level traffic side channel that can expose model, prompt, and task behavior despite encryption. From packet sizes, directions, timing, and burst structure alone, a passive local observer can infer the serving model, the user's prompt category, and the task executed by a collaborative multi-agent system. Yet current evidence is fragmented across separate datasets and settings, limiting reproducibility and comparison. We present, to our knowledge, the first unified measurement study and public benchmark of encrypted LLM traffic across both user--LLM and multi-agent executions. The large-scale benchmark contains 60,000 user--LLM interactions across 10 models and 6 prompt categories, plus