Your organization uses Microsoft Purview to classify data. You need to automatically apply a 'Confidential' label to documents that contain salary information. Which type of sensitivity label configuration should you use?
Auto-labeling policies in Microsoft Purview are designed to automatically detect specific content within documents and emails, such as financial data or personally identifiable information, by using sensitive information types (SITs). When a defined SIT or condition is met, the policy automatically applies the corresponding sensitivity label without requiring any user interaction. This method is ideal for consistent, content-based data classification across an organization.
Why this answer
Auto-labeling with sensitive information types allows Microsoft Purview to automatically apply a 'Confidential' label to documents containing salary information based on predefined patterns, such as employee salary data. This configuration uses content scanning to detect sensitive data (e.g., via built-in or custom sensitive information types) and applies the label without user intervention, meeting the requirement for automatic classification.
Exam trap
The trap here is confusing auto-labeling with default labeling or mandatory labeling policies, as candidates may think a default label or a policy requiring user input can achieve automatic content-based labeling, but only auto-labeling with sensitive information types performs dynamic content scanning.
How to eliminate wrong answers
Option A is wrong because manual labeling requires users to apply labels themselves, which does not meet the requirement for automatic application. Option C is wrong because default labeling for SharePoint libraries applies a label to all documents in a library by default, but it does not dynamically detect salary information; it applies a static label regardless of content. Option D is wrong because mandatory labeling policy requires users to select a label when saving or sharing documents, but it does not automatically apply a specific label based on content detection.