Which THREE are valid methods to protect sensitive data in BigQuery?
Trap 1: Create authorized views that query only non-sensitive columns.
Incorrect. Authorized views restrict access to rows or columns but do not encrypt or de-identify data; they are an access control mechanism, not a direct data protection method.
Trap 2: Use IAM roles to grant access at the dataset level, which…
Incorrect. IAM roles grant access at the dataset level but do not automatically mask sensitive data; they control overall access, not data masking.
- A
Enable customer-managed encryption keys (CMEK) to encrypt sensitive columns.
Correct. CMEK encrypts data at rest, protecting sensitive columns from unauthorized access to storage layers.
- B
Apply Cloud DLP de-identification transforms during data ingestion.
Correct. Cloud DLP de-identification transforms can mask, tokenize, or redact sensitive data before it is stored in BigQuery.
- C
Create authorized views that query only non-sensitive columns.
Why wrong: Incorrect. Authorized views restrict access to rows or columns but do not encrypt or de-identify data; they are an access control mechanism, not a direct data protection method.
- D
Use BigQuery column-level security to restrict access to sensitive columns.
Correct. Column-level security (Policy Tags) restricts access to sensitive columns by authorized users only.
- E
Use IAM roles to grant access at the dataset level, which automatically masks sensitive data.
Why wrong: Incorrect. IAM roles grant access at the dataset level but do not automatically mask sensitive data; they control overall access, not data masking.