← Back to glossary

Parameter controlling cumulative privacy loss in differentially private systems.

Definition

A privacy budget (often denoted epsilon, δ) measures and limits the cumulative privacy loss from multiple analyses. Systems track budget usage and restrict further queries when exhausted.

In plain English Parameter controlling cumulative privacy loss in differentially private systems.

Why this matters

Why it matters: It prevents “death by a thousand cuts” where many small releases reveal sensitive facts.

Example

Example: DP analytics platform allocates budget per dataset and per analyst role.