LATIDIA · Ciberseguridad

Comprender y mitigar la suplantación de roles en los LLM

arXiv: 2610.40396v1Tipo de anuncio: nuevo Resumen: A pesar de las crecientes capacidades, los LLM siguen siendo susceptibles a inyecciones rápidas: entradas maliciosas que pueden manipular o secuestrar modelos. En particular, recientemente ha sido observador

La noticia

arXiv:2610.40396v1 Announce Type: new Abstract: Despite growing capabilities, LLMs remain susceptible to prompt injections: malicious inputs that can manipulate or hijack models. In particular, it has recently been observed that LLMs can exhibit role confusion (Ye et al., 2026), where models attribute the source of text based on its style instead of its labeled role. Malicious text in tool outputs (for instance, a malicious webpage accessed by an agent) can leverage role confusion to impersonate the user, an attack we call role spoofing}. In this work, we seek to understand and mitigate role spoofing. We study how models track "who said what" in a conversation, showing that they use cues