Linkage risk is the chance that separate data fields, which may look harmless on their own, can be combined to identify a person. A patient ID, zip code, device token, or contact list may not directly name anyone, but when matched with another dataset it can reveal identity, health status, or other sensitive details. This is especially important in healthcare, where coded or pseudonymous records are often still protected if they can be traced back to an individual.
In cyber security, linkage risk matters because attackers do not need a full record to cause harm. They can correlate partial data across breached systems, partner databases, public sources, or leaked directories to rebuild identities. Defenders reduce this risk by minimizing shared identifiers, separating sensitive systems, enforcing least privilege, and monitoring access to fields that enable re-identification. Good privacy design is not only about hiding names; it is about preventing data from being stitched together after the fact.