How We Can Save Anonymization
summary
Argues that effective anonymization exists and gives concrete checks: skepticism, rigorous threat models that include grad students and intimate‑partner attackers, avoidance of fragile schemes that rely on secrecy or governance, and preference for composability guarantees so releases degrade gradually. Cites a successful Wikipedia browsing release and Census 2020 release as examples, warns about catastrophic failures like combining ZIP‑keyed hospital data with cancer slices or naive k‑anonymity k=101, and insists methods must be explainable and peer‑reviewable.
AI-written summary · about
chapters (9)
- 0:00Anonymization in peril
- 1:26Why anonymization fails
- 3:11Industry standards lacking
- 5:19Building anonymization standards
- 6:01Evaluating anonymization techniques
- 10:14Governance vs anonymization
- 12:01Avoiding anonymization failures
- 13:25Transparency for trust
- 14:09Path forward for anonymization
related talks
01h 17mTo be or not to be (anonymous)? Anonymyity in the age of big and open data.
32mThe Dangerous Data Anonymization - Keynote
36mWhat every data scientist should know about data anonymization
29mPersonal Data Pseudonymization Versus Anonymization In The Age Of AI & Big Data
33mPractical Data Privacy
10mHealth Data Anonymization and Privacy
01h 19mAcademic sessions 1.
50mPrivacy in Practice with Smart Pseudonymization: Lessons from the Belgian Public Sector