LATIDIA · Ciberseguridad

Más allá de la inyección de mensajes: la brecha en la actualización de herramientas en la seguridad del agente LLM

arXiv:2611.01976v1 Anuncio Tipo: nuevo Resumen: Los agentes basados ​​en modelos de lenguaje grandes dependen cada vez más de herramientas externas para invocar servicios e interactuar con recursos web. Estas herramientas pueden evolucionar a través de actualizaciones de software, y

La noticia

arXiv:2611.01976v1 Announce Type: new Abstract: Large language model based agents increasingly rely on external tools to invoke services and interact with web resources. These tools can evolve through software updates, yet agentic frameworks may continue trusting them without re-verifying their safety after implementation changes. This tool update gap creates a supply-chain vulnerability in agentic systems: because updated tool code executes outside the LLM's context window, malicious functionality embedded in tool updates can bypass defenses centered on model-visible interactions, leaving agents exposed to stealthy, persistent threats that existing in-context defenses can hardly address. To thoroughly investigate this threat, we introduce ETUDE (Exploiting Tool Updates for Direct Execution), an attack that embeds