Predefined masking policies

Use the predefined masking policy reference topics to understand the behavior, input requirements, configuration options, and known limitations for each predefined masking policy available in IBM® Optim Data Privacy.

Predefined masking policies are a separate category from standard masking policies. Unlike standard policies, which use standard format names and processor types, the predefined policies are configured directly through the IBM Optim Data Privacy UI or through the policy_metadata and processor blocks in the API. Each predefined policy targets a specific sensitive data type and exposes its own set of configuration options.

The following predefined masking policy types are available:

Common behavior across all predefined policies

The following behavior applies to all predefined masking policies:

  • A NULL input always produces a NULL output. No masked value is generated and no error is raised.

Repeatable and random algorithms

Most predefined masking policies support two masking algorithms that control whether the output is deterministic or non-deterministic.

Repeatable
Produces a deterministic output. The same input value always produces the same masked output, regardless of when or how many times the job runs. Use the repeatable algorithm for columns where referential integrity must be maintained, such as when the same value appears in multiple tables and must consistently map to the same masked value.
Random
Produces a non-deterministic output. Each transformation generates a new value, even when the input is identical to a previous row. Use the random algorithm for columns where no correlation between rows is required and run-to-run consistency is not needed.