Courts analytics must let junior analysts query docket tables while redacting defendant dates of birth. Which Unity Catalog control should they configure?

Applying a column mask that redacts defendant dates of birth for junior analyst roles is the correct Unity Catalog control for courts analytics. Column masks let junior analysts query docket tables for legitimate case-load work while transforming or nulling the DOB column for those principals, preserving a single governed docket table instead of fracturing storage by role. Applying a row filter that masks defendant dates of birth for junior analyst roles confuses row-level control with column-level control because row filters hide entire rows rather than masking values in the DOB column, so juniors would lose needed docket records while the column itself remains unmasked on visible rows. Applying a table grant that masks defendant dates of birth for junior analyst roles confuses table-level access with column-level masking because grants only allow or deny SELECT on the whole table rather than transforming values in the DOB column, so juniors granted access would still see unmasked DOB values. Creating a dynamic view that redacts defendant dates of birth for junior analyst roles requires maintaining a separate object with duplicated logic rather than applying a centralized column mask on the governed docket table, losing role-based enforcement and single-table governance. Exam caveat: Column masks differ from denying SELECT on the whole table and from row filters. Juniors still need table access for non-sensitive docket attributes; masks target specific columns by principal or role. Masks also are not a substitute for encryption at rest or for audit logging of who queried court PII. Operational check: Configure the mask on the defendant DOB column for junior analyst roles, query as a junior and confirm DOB is redacted while other docket fields remain usable, query as a privileged role that should see DOB and confirm unmasked values, and verify juniors lack MODIFY on the production docket tables.