AIExplainer

What is differential privacy?

A method to protect sensitive information in datasets by adding noise to the data

Differential privacy is a way to keep personal data private when it's being used for research or analysis. It works by adding random noise to the data, making it hard for others to figure out individual information

Think of differential privacy like a secure voting booth. Just as you can't see who someone voted for, differential privacy makes it impossible to identify individual data points in a dataset, while still allowing researchers to see overall trends

The US Census Bureau uses differential privacy to protect the personal data of citizens while still providing accurate population statistics

Differential privacy is used in data analysis, machine learning, and research to protect sensitive information, such as medical records or financial data

Some people think differential privacy makes data completely anonymous, but it's not. It just makes it very hard to identify individual data points

Differential privacy was first introduced in 2006 by researchers Cynthia Dwork, Frank McSherry, Kobbi Nissim, and Adam Smith

data privacy statistical privacy privacy-preserving data analysis

Three products for different needs — explore what’s relevant to you.