LATIDIA · Ciberseguridad

Renunciar a la privacidad

arXiv: 2609.10992v2Tipo de anuncio: replace-cross Resumen: La integración de los modelos de lenguaje grandes en las tareas diarias se basa en instrucciones ricas en contexto, exponiendo inevitablemente información confidencial del usuario. Privacidad actual-pre

La noticia

arXiv:2609.10992v2 Announce Type: replace-cross Abstract: The integration of Large Language Models into daily tasks relies on context-rich instructions, inevitably exposing sensitive user information. Current privacy-preserving methods typically employ context-agnostic static rules, causing severe utility degradation. However, the specific mechanisms governing how sanitization impacts downstream performance remain largely underexplored. To address this, we conduct a systematic analysis to deconstruct the privacy-utility trade-off, uncovering three underlying mechanisms: (1) Context-Dependent Utility, which first establishes when to sanitize by revealing that data value shifts from critical constraints to dispensable noise based on user intent; (2) Strategic Adaptation, which subsequently determines how to sanitize by dictating that the choice between removal and replacement depends on