LATIDIA · Ciberseguridad

Más allá de la formación privada: el nuevo panorama de la privacidad de la IA

arXiv: 2609.19456v1Announce Type: new Resumen: Los sistemas aumentados por recuperación dependen cada vez más de índices vectoriales que pueden retener elementos eliminados en su gráfico de búsqueda. Las interfaces de eliminación existentes pueden evitar que se elimine la identidad

La noticia

arXiv:2609.19456v1 Announce Type: new Abstract: Retrieval-augmented systems increasingly rely on vector indexes that may retain deleted items in their search graph. Existing deletion interfaces can prevent deleted identifiers from appearing in returned results while still computing distances to their embeddings during graph traversal. We formalize this distinction as output safety versus traversal safety, and introduce TSD-AUDIT, a framework for auditing and enforcing traversal-safe deletion in graph-based approximate nearest-neighbor retrieval. On Faiss IndexHNSWFlat, native filtering leaves the number of distance computations unchanged relative to unfiltered search; at a 70% deletion rate, trace-faithful replay detects deleted-vector scoring in all 100 audited queries. Code inspection of hnswlib's mark_deleted path reveals the same scoring-before-liveness