Azure AI Search adds Microsoft Purview sensitivity label support to knowledge sources in public preview. Labels applied at the source system flow through ingestion into Azure AI Search and into the knowledge bases that ground Foundry agents and copilots, so chunk retrieval, vector search, and multi-step agent reasoning all respect the same access and privacy classifications defined by enterprise security teams. Customers building secure agentic workflows over regulated data get end-to-end label fidelity instead of having to rebuild access enforcement at the retrieval layer. The capability is a prerequisite for downstream Purview-aware agent orchestration scenarios. Existing knowledge sources continue to work; sensitivity awareness turns on per source. The release covers indexed Purview-labeled content surfaced through the Knowledge Source API.
Learn more.