LATIDIA · Ciberseguridad

Las capacidades del agente de LLM deben seguir la intención de la tarea y la fuente del contexto

arXiv:2609.14631v1 Tipo de anuncio: nuevo Resumen: los agentes de LLM realizan acciones reales, que incluyen ejecutar código, modificar archivos, llamar a servicios y delegar tareas, impulsadas por fuentes de contexto: solicitudes de usuario, resultados de herramientas, documentos

La noticia

arXiv:2609.14631v1 Announce Type: new Abstract: LLM agents take real actions, including executing code, modifying files, calling services, and delegating tasks, driven by context sources: user requests, tool results, documents, shell outputs, Skill and MCP instructions, memory. Unlike traditional systems, where capability is predefined, the least-privilege capability an agent needs is dynamic, depending on its task intent: what it wants to do and how. This creates a security and safety challenge: all inputs enter one shared planning channel with equal influence, by adversarial injection or accidental scope widening, yet a user's explicit request and a document's extracted text carry different trust and should not share authority to choose a destination or