LATIDIA · Robótica

Entornos de pruebas controlados de IA: un modelo de amenazas, una taxonomía y un marco de medición

arXiv:2606.18532v2 Tipo de anuncio: sustitución cruzada Resumen: Los entornos de pruebas controlados de IA generan evidencias sobre sistemas que pueden actuar a través de herramientas de software, dispositivos conectados en red o actuadores físicos. La interpretación de esas pruebas dep

La noticia

arXiv:2606.18532v2 Announce Type: replace-cross Abstract: AI sandboxes produce evidence about systems that may act through software tools, networked devices, or physical actuators. The interpretation of that evidence depends on the boundary of the experiment. We define a sandbox apparatus, distinguish five overlapping archetypes, and specify fifteen claim-relative measurement dimensions. A weakest-link licensing rule caps a deployment claim at its weakest necessary evidence dimension. We connect this rule to a threat model covering both outward harm and corruption of the evaluation apparatus, then apply it to four worked examples spanning driving, industrial control, tool-using agents, and vision-language-action manipulation. We also map the required artifacts to standards and regulatory instruments. The framework