GCP Professional Data Engineer Practice Question

Creating distinct Google Cloud projects for development and production gives each environment its own IAM policy, quotas, and billing, guaranteeing that an error in development cannot affect production data. Running each Dataflow pipeline with a dedicated service account that has only the permissions it needs enforces least privilege. In production, exposing the data to developers through an authorized BigQuery view that omits or masks sensitive columns lets them work with representative data while preventing direct access to PII. Granting developers only the BigQuery Data Viewer role on that view, and higher-level roles (for example, BigQuery Data Editor) only in the development project, limits the attack surface and administrative overhead. The other choices leave both environments in a single project or continue to grant excessive permissions, so they fail to isolate production or to adequately protect sensitive data, or they introduce unnecessary data-copy operations that increase operational complexity.