Hybrid Legal-Technical Concepts of Privacy
Talkat Data Privacy: From Foundations to Applications (Simons Institute workshop, 2019) · 2019-03-04..03-08·
summary
Argues that common legal concepts like HIPAA safe harbor fail empirically, builds a practical catalog of controls and a tiered access architecture that uses differential privacy for open queries, sanitized downloads under agreements, and secure enclaves with audit logs for high‑risk data. Extracts a formal, game‑based model of FERPA (adversary given directory info) and proves mathematically that a DP mechanism meets that conservative legal requirement. Proposes formalizing hybrid concepts such as singling out so regulations state measurable privacy goals instead of ad hoc deidentification.
AI-written summary · about
chapters (16)
- 0:00Intro
- 1:06Inadequacy of Current Regulatory Framework
- 3:27Weaknesses of Common Concepts
- 5:09The Inherent Hybrid Nature of Data Privacy
- 6:55Gaps between Technical & Normative Concepts
- 23:12Framework for Privacy-Aware Data Releases
- 28:55Approach to Formal Modeling
- 29:14Illustration: Modeling FERPA
- 30:00Modeling FERPA: The Adversary
- 30:39Modeling FERPA: Directory Information
- 31:27Components of a FERPA Privacy Game
- 32:45Common Privacy Concepts in the Law
- 33:02Interpreting the Differential Privacy Guarantee
- 33:35Personally Identifiable Information
- 34:28PII: Interpretation of DP Guarantee
- 36:10Introducing Hybrid Concepts
related talks
01h 16mLaw and CS
29mKeynote Presentation: Minding the (Privacy Research/Practice) Gap
01h 14mPerspective: Achieving Meaningful Privacy in Technology and Design
29mPrivacy-preserving methods: Building secure projects
09mIntegrating Differential Privacy and Contextual Integrity
01h 10mDIGITAL PLATFORMS, THE NEW PRIVACY CHAMPIONS? BETWEEN MYTHS AND REALITIES
01h 00mPANEL: Statistical Theory, Privacy and Data Analysis
10mPrivacy-Preserving Analytics on the Ground