Insights without exposure

How should fraud examiners and legal professionals meet compliance standards but also keep individuals’ and organizations’ data private? “Differential privacy” is a system of cybersecurity that proponents claim can protect data far better than traditional sanitizing or anonymizing methods.

Over the past two years as a columnist for Fraud Magazine, it’s been a pleasure introducing new ideas, innovations and technologies to you — my colleagues. That’s why I’m excited to devote this edition of “Innovation Update” to the concept of “differential privacy.”

Differential privacy can securely limit algorithms so organizations can securely share private, sensitive data internally or among third parties. The concept isn’t new. Mathematicians, cryptographers and academics have been discussing it for more than a decade. However, companies are now commercializing it for global fraud examinations and proactive compliance monitoring.

In September 2019, Google released the open-source version of the differential privacy library it uses in some of its products, such as Maps, according to Emil Protalinski, the author of Google open-sources its differential privacy library, VentureBeat, Sept. 5, 2019.