LATIDIA · Ciberseguridad

Inyección rápida a través de evidencia forense: una taxonomía y prueba de concepto de ataques antiforenses en investigaciones digitales asistidas por LLM

arXiv: 2610.29747v1Tipo de anuncio: nuevo Resumen: los modelos de lenguaje grandes (LLM) están ingresando a la ciencia forense digital como asistentes de examen y agentes autónomos. El análisis forense digital invierte la configuración habitual para el aviso indirecto

La noticia

arXiv:2610.29747v1 Announce Type: new Abstract: Large language models (LLMs) are entering digital forensics as examination assistants and autonomous agents. Digital forensics inverts the usual setting for indirect prompt injection: the evidence under analysis is attacker-controlled by definition, and the examiner is obliged to read it closely. This makes the reasoning layer a new anti-forensic target, next to the classical ones of data and tools. Classical anti-forensic taxonomies do not cover it, and prompt injection defences do not anticipate it. We extend the anti-forensic model to reflect the manipulation of reasoning, and present a taxonomy of attacks carried through adversarial evidence: an attacker selects an objective, a technique, a payload location,