Bigquery Authorized View Secure Data Sharing

What BigQuery Authorized Views Are

An authorized view in BigQuery is a logical SQL view that allows specific users or groups to query aggregated or filtered data from a source dataset without granting those users direct access to the underlying source tables. By authorizing the view itself to access the source dataset, BigQuery executes queries against the view using the view's elevated dataset authorization rather than the end-user's permissions.

How It Meets the Requirements

• Pre-aggregated metrics on demand: The SQL view definition performs the required aggregation transformations dynamically when queried.
• Prevent access to raw records: The partner is only granted access to the separate dataset containing the view. Because the view exposes only aggregated outputs, individual transactional PII remains inaccessible.
• No permissions on source dataset: Because the view itself is explicitly added to the authorized views list of raw_transactions, users only need the roles/bigquery.dataViewer role on the dataset containing the view.
• Audit trails and compliance: Enabling BigQuery Data Access audit logs (specifically DATA_READ logs) in Cloud Audit Logs ensures every query run against the authorized view is captured with detailed metadata, timestamps, and caller identities.

Key Advantages

• Least Privilege Access: Eliminates the need to grant table-level or dataset-level read permissions on sensitive source data.
• Zero Data Duplication: Eliminates ETL pipeline overhead and storage redundancy by executing queries on demand.
• Centralized Security: Revoking access to the view dataset immediately terminates external partner access.
• Granular Governance: Works seamlessly with IAM roles like roles/bigquery.dataViewer and Cloud Logging.

Why This Is Optimal

This pattern separates consumer presentation from protected underlying storage while maintaining complete regulatory auditing and strict data isolation.