Secure the storage data plane behind AI retrieval
Separate ingestion and serving authority, close Shared Key and anonymous paths, constrain SAS use, and validate Blob and Data Lake access independently.
Explore the topic
Distinguish Entra authorization, shared keys, SAS and anonymous access in the data plane.
Questions this collection helps you answer
Build your understanding
Separate ingestion and serving authority, close Shared Key and anonymous paths, constrain SAS use, and validate Blob and Data Lake access independently.
Choose between application security filters and native preview permissions, preserve access metadata on chunks, and enforce the same boundary across keyword and vector retrieval.
Move from control to evidence
Use these as investigation leads. Table coverage and context determine what the results can establish.
Inventory successful storage-account and private-endpoint connection mutations, then compare configuration evidence before claiming that network access was widened.
Identify tenant-qualified OAuth principals whose completed daily GetBlob volume exceeds a fixed historical average, retaining response-byte and object-diversity context.
Quantify AccountKey-authenticated Blob requests by resource, client and operation category while retaining outcomes and transfer context.
Review SAS request outcomes and expiry-policy evidence without exposing token-bearing URLs or assuming all SAS access uses the same signing authority.
Group successful DeleteBlob requests by account and caller evidence, retaining object samples while distinguishing request counts from permanent data destruction.
Separate successful anonymous reads from rejected requests, with sanitized object paths and response-size context rather than assuming every anonymous hit exposed data.
Follow the connections