LATIDIA · Ciberseguridad

Colapso de seguridad interna en LLM y agentes de codificación

arXiv: 2603.23509v2Announce Type: replace-cross Resumen: Los LLM y los agentes de codificación ahora pueden abordar tareas complejas y de largo horizonte a través de avances en el aprendizaje de refuerzo, pero no es necesario contar con capacidades de resolución de tareas más sólidas

La noticia

arXiv:2603.23509v2 Announce Type: replace-cross Abstract: LLMs and coding agents can now tackle complex, long-horizon tasks through advances in reinforcement learning, but stronger task-solving capabilities do not necessarily preserve safety alignment. Recent AI agent incidents involving reward hacking indicate how optimizing for task success can bring learned capabilities into conflict with safety alignment. In this work, we identify a critical failure mode that manifests this conflict, which we term Internal Safety Collapse (ISC). When completing otherwise benign coding tasks, AI agents can prioritize satisfying program requirements over safety constraints, generating large volumes of sensitive content, including unsafe outputs, as task data. To study this phenomenon, we introduce TVD (Task, Validator, Data),