What is the Origin of Data Linkage Attacks?

This term originates from the field of data privacy, computer science, and health informatics, emerging as a critical concern with the rise of big data and the sharing of large, ostensibly anonymized datasets for research and commercial purposes. The vulnerability was scientifically demonstrated through various studies showing the high probability of re-identification even with robust de-identification techniques. It is a concept central to discussions on HIPAA compliance and genetic data security.

What is the Mechanism of Data Linkage Attacks?

The attack operates by identifying quasi-identifiers—non-unique pieces of information that, when combined, become unique to an individual, such as birth date, zip code, and gender. The attacker cross-references these quasi-identifiers across two or more separate databases, which may include clinical data, consumer purchase records, or public demographic registries. The convergence of these data points allows for the high-confidence deduction of the individual's identity, effectively defeating the initial de-identification efforts.

Data Linkage Attacks

Meaning

A cybersecurity threat model where an attacker combines or “links” seemingly anonymized or de-identified datasets from multiple sources to re-identify individuals and infer sensitive personal information. In the health domain, this attack can connect de-identified medical records with publicly available or other non-medical data, compromising patient privacy and confidentiality. This method exploits the uniqueness of individual data points across disparate databases to reconstruct a personal identity.