I don't need privacy, I got confidentiality!
summary
Claim: privacy differs from confidentiality and must be engineered into the SDLC using concrete techniques. Uses threat modeling (LINDDUN-style categories) to find privacy risks, cites concrete failures: Roomba image leaks, smart‑meter TV inference, menstrual‑app exposure after Roe v Wade, and a Belgian DPA €25,000 daily fine for dark‑pattern consent. Recommends minimization, data inventories, consent management, deidentification (differential privacy, synthetic data with caveats) and integrating privacy controls into design and testing.
AI-written summary · about
chapters (16)
- 0:00Introduction
- 4:32Threat modeling basics
- 9:38Why privacy matters
- 10:42Security vs privacy triads
- 15:00Dark patterns in compliance
- 17:53Cars collecting personal data
- 21:08Privacy by design solutions
- 24:44Security vs privacy conflicts
- 27:35Security vs privacy mindset
- 30:26Integrating privacy into security
- 33:12Threat modeling foundation
- 35:24Reasonable privacy expectations
- 38:40Starting small with privacy
- 39:42Leveraging security for privacy
- 42:33Summary and call to action
- 42:55Q&A
related talks
44mA Taste of Privacy Threat Modeling
50mShifting Privacy In: How Privacy and Security Can Strengthen Each Other
01h 15mDATA PROTECTION ENGINEERING: WHAT IS THE ROAD AHEAD?
08mPrivacy by Design for DevOps
29mPrivacy-preserving methods: Building secure projects
40mData Privacy and Security by Design
43mPanel: Privacy Design Patterns for AI Systems: Threats and Protections
52mDeveloping privacy