Dataset linking is the process of combining two or more data sources so that records can be matched to the same person, device, or account. The match may rely on direct identifiers, but it often uses indirect signals such as location, timestamps, job role, purchase history, or other quasi-identifiers. In privacy and security work, linking is important because data that looks harmless on its own can become identifying when it is joined with other records.
In cyber security, dataset linking is used in both attacks and defenses. Attackers may correlate breached data, public profiles, and telemetry to uncover identities, habits, or sensitive relationships, even when names were removed. Defenders use controlled linkage to detect fraud, join threat intelligence, and validate whether anonymization is still credible. The main risk is re-identification: once multiple datasets are combined, the original protection may fail if the overlap is enough to single out a person or reveal a pattern. Strong defenses include data minimization, access control, careful sampling, and repeated testing of how easily records can be linked.