Evaluating the Usability of Differential Privacy Tools with Data Practitioners

Explore a 13-minute conference talk from SOUPS 2024 examining how data practitioners interact with differential privacy (DP) tools. Learn about a comprehensive usability study conducted with 24 US data practitioners evaluating four open-source Python-based DP tools: DiffPrivLib, Tumult Analytics, PipelineDP, and OpenDP. Discover key findings on how these tools support practitioners' understanding and implementation of differential privacy, with particular focus on the critical role of API design and documentation. Gain insights into evidence-based recommendations for improving DP tool usability to encourage wider adoption of differential privacy in real-world data analytics applications.