How data masking works and what data is masked

When data masking is enabled, certain customer identifiers are handled automatically, while others require you to explicitly mark them as personally sensitive data. Once a field is marked, the same masking behaviour is applied consistently across Member Care, API responses, and Databricks.

Data masking in Member Care & APIs

Data type	Masked by default	Requires explicit marking
Core customer identifiers (first name, last name, and mobile, email)	Yes	No
Identifiers under communication channel	Yes	No
Customer profile custom fields	No	Yes (mark as PSI)
Customer profile extended fields	No	Yes (mark as PSI)
Behavioural event attributes	No	Yes (mark as PSI)

📘
Note
External IDs are not masked. Card data, Transaction or billing data, and other data other than the mentioned above data are not masked.

Data masking in Databricks

In Databricks, sensitive customer data is masked by default and displayed as asterisks (*****). Unmasked values are shown only when the user or job has been explicitly authorised to access sensitive data.

Data category	Masked in Databricks	Notes
Core customer identifiers	Yes	First name, last name, email, mobile number, external ID
Customer profile custom fields	Yes, if marked as PSI	Must be explicitly marked to be masked
Customer profile extended fields	Yes, if marked as PSI	Must be explicitly marked to be masked
User (customer) profile tables	Yes	Masking applies to PSI-marked fields
Type-2 (history-tracking) user dimension tables	Yes	Applicable only to the USHC cluster
Card data	No	Not supported
Transaction or billing data	No	Not supported
Behavioural event data	No	Not masked in Databricks
Any other non-customer data	No	Not supported