Protecting data requires matching the right control to the right data state and context. The CISSP exam tests three categories of data protection control that operate across different dimensions: Data Loss Prevention (DLP) detects and prevents unauthorised data movement, Digital Rights Management (DRM) restricts what authorised users can do with data, and Cloud Access Security Brokers (CASB) extend visibility and control to cloud-hosted data. Understanding how these controls differ, where they apply, and when to recommend each is essential for Domain 2.

Data States and Appropriate Controls

Before selecting data protection controls, security professionals must understand the three states in which data exists and the distinct security challenges each presents.

Data at rest is data stored on a persistent medium: a hard drive, SSD, database, backup tape, or cloud storage bucket. The primary threat to data at rest is unauthorised access — either through physical theft of the media or logical access by an attacker who has compromised the storage system. Controls for data at rest include encryption (AES-256 is the standard), access controls (filesystem permissions, database access controls, cloud IAM policies), and physical security for storage media.

Data in transit is data moving between systems, users, or organisations across a network. The primary threats are interception (eavesdropping) and modification (man-in-the-middle attacks). Controls for data in transit include transport encryption (TLS 1.3, IPSec), message-level encryption (S/MIME, PGP for email), secure transfer protocols (SFTP, HTTPS), and digital signatures to verify integrity.

Data in use is data being actively processed in memory, displayed on screen, or manipulated by a user or application. This is the most difficult state to protect because encryption must be removed for the data to be processed. Controls for data in use include access controls on applications, session management, screen capture prevention, and emerging technologies like confidential computing (which performs calculations on encrypted data without decrypting it in unprotected memory).