LATIDIA · Ciberseguridad

Privacidad contrastiva: un enfoque semántico para medir la privacidad de la desinfección basada en IA

arXiv: 2605.02977v2Tipo de anuncio: reemplazar Resumen: La desinfección basada en IA puede eliminar conceptos de imágenes y texto, pero la evaluación de la privacidad sigue siendo en gran medida ad hoc. Proponemos la privacidad contrastiva, una definición formal que

La noticia

arXiv:2605.02977v2 Announce Type: replace Abstract: AI-based sanitization can remove concepts from images and text, but privacy evaluation remains largely ad hoc. We propose contrastive privacy, a formal definition that yields a quantitative test with a semantic interpretation. Under formal assumptions, we derive a conditional sufficiency result for a class of sanitized renderings (i.e., media files). We operationalize the definition using imperfect semantic-distance models such as CLIP. The test compares sanitized renderings under audit with both the original and sanitized versions of reference renderings known to contain privacy-relevant properties; if the rendering under audit is semantically closer to the unsanitized reference, then the former might leak private information even after sanitization.