Create first: Before creating higher-level data protection constructs (like Data Loss Prevention policies or sensitivity labels) that rely on custom definitions, you must first create A sensitive info type (SIT). The SIT defines the unique pattern or structure of the data you want to identify and protect.
Use for detection method: To accurately capture strings or values that follow a rigid, predictable format (such as employee numbers, specific ID formats, or customized serial codes), a Regular expression (regex) is the ideal detection mechanism. It offers the precision required to match specific patterns rather than relying on exact word matches (Keywords) or static lists (Dictionary).
| Admin | Role Assigned | Can Create Sensitivity Labels? |
| Admin1 | Global Reader | No, read-only permissions. |
| Admin2 | Compliance Data Administrator | Yes, can manage compliance data, including labels. |
| Admin3 | Compliance Administrator | Yes, has full compliance management, including labels. |
| Admin4 | Security Operator | No, this role is focused on security alerts and response. |
| Admin5 | Security Administrator | No, primarily focused on security policies and threat management. |
Step 1: Create a sensitive info type. Before a custom category of data can be classified automatically, you must first define the criteria used to identify it (such as a specific pattern or regex). Creating a sensitive info type forms the building block for detection.
Step 2: Create a sensitivity label. Once the detection criteria exist, you must create the container that defines what happens to that data (e.g., applying encryption or content markings) when it is identified.
Step 3: Create an auto-labeling policy. This policy brings everything together by scanning content locations (such as SharePoint, OneDrive, or Exchange) for the sensitive info type defined in Step 1 and automatically applying the sensitivity label created in Step 2.