No free lunch in data privacy

D Kifer and A Machanavajjhala


A foundational paper on data privacy, that has spawned much follow-up work, and belongs in the curriculum of any course on data privacy. It raises fundamental questions on how to define privacy, and the situations when differential private mechanisms provide meaningful semantic privacy guarantees. It led to the development of more practical, general privacy frameworks. Seminal ideas from the paper are being used to quantify the privacy ensured by the 2020 Decennial Census to each person in the US.