LATIDIA · Ciberseguridad

ActGov: Gobierno de las acciones de los agentes de LLM a través de la validación restringida por políticas

arXiv: 2609.24446v2 Tipo de anuncio: reemplazar Resumen: Los agentes del modelo de lenguaje grande (LLM) ejecutan cada vez más flujos de trabajo de largo horizonte a través de herramientas externas, lo que permite que las salidas no confiables influyan en las acciones posteriores y exc

La noticia

arXiv:2609.24446v2 Announce Type: replace Abstract: Large language model (LLM) agents increasingly execute long-horizon workflows through external tools, allowing untrusted outputs to influence subsequent actions and exceed user authorization. Existing defenses isolate injected content or constrain execution with predefined plans and static policies, but these approaches are brittle under dynamic workflows and scale poorly across extensible tool ecosystems. In this work, we present ActGov, a runtime enforcement framework that validates each LLM-proposed tool action before it causes external effects. Built on a unified semantic model of authorization, actions, runtime context, and security constraints, the ActGov-Policy component iteratively constructs a policy set from tool specifications, benign tasks, and observed failure traces, with