Deep Dive Workshop – Data Security and Data Protection
summary
Cloud use is now essential but comes with concrete trade-offs: calibrate security, privacy and performance; run DPIAs and transfer impact assessments after Schrems II; expect €4.3 million average loss per cyber incident. Access control and identity directories (Microsoft Azure AD, AWS/GCP tooling) are immediate wins; cryptography protects data at rest and in transit but not while processing, so apply function separation: assignment rule (H) / processing (P) / calculation (C) using HSMs to erase mappings. Deploy differential privacy (open-source libs, TensorFlow Privacy) for safe analytics (used in Maps and COVID mobility reports) and adopt real‑time pseudonymization (Data Vaccinator) to minimize breach impact.
AI-written summary · about
chapters (21)
- 0:00Start
- 5:11Privacy and Security of Zoom
- 8:58Emerging Laws
- 11:42Data Protection Impact Assessments
- 12:09Technical Considerations
- 12:43Governance
- 16:14International Transfers
- 21:17The Discovery of Data
- 30:26Risks of Data Leakage at Home
- 39:45Security of Processing
- 43:50End-to-End Encryption
- 45:13Encryption
- 48:50Microsoft Azure
- 1:05:01Differential Privacy
- 1:05:50Example To Illustrate What Differential Privacy Is
- 1:14:32Differentially Private Libraries
- 1:15:11Machine Learning
- 1:15:35Tensorflow Privacy
- 1:20:31How Do You Remove the Noise To Read the Accurate Data
- 1:21:31Can You Apply Differential Privacy to Um Geographical Tracking Data
- 1:38:19Pseudonymization
related talks
39mData Privacy Engineering to power Responsible Enterprise AI
26mAnalytics & Data Science with Privacy at its Core
01h 15mDATA PROTECTION ENGINEERING: WHAT IS THE ROAD AHEAD?
39mdev.privacy: GDPR in a nutshell
08mPrivacy by Design for DevOps
52mDeveloping privacy
01h 20mUnlocking societal benefits of AI with privacy protective technologies.
32mNew Developments in Cloud Technology and Data Security For Practitioners