← AI Glossary

t-Closeness

A privacy model requiring that the distribution of a sensitive attribute within any equivalence class is close to the distribution in the overall dataset, measured by the Earth Mover's Distance.

In plain language

A privacy protection method that prevents attackers from narrowing down a person's private details. It ensures that if you look at any group in the data, the patterns of sensitive information look similar to the overall population.

Why this matters

T-closeness is a technical control in the privacy-by-design toolkit. It strengthens data anonymisation beyond simpler methods, reducing inference risk in datasets. For organisations managing sensitive information under the Privacy Act, t-closeness provides a defensible technical standard for data de-identification.

Relevance

Implementation

Offers a formal privacy model to protect against attribute inference attacks, supporting defensible data anonymisation in regulated environments.

Putting t-closeness into practice in your organisation?

Ready to transform your AI strategy?

Partner with Australia's AI strategy and governance specialists. From adoption roadmaps to ISO 42001 audit readiness.