Q&A with Keynote – Marit Hansen
summary
Answers focus on concrete rules and fixes: GDPR allows processing under Article 6 with mitigations like encryption or pseudonymization, but consent blocks profiling; logkeeping requirements under Article 30 force documentation and risk assessments. Practical solutions include storing aggregated or hashed identifiers off‑chain, trustees for key management, DPIAs for high‑risk ML, and producing math/statistical audit reports rather than raw model code. Regulators and cross‑border enforcement (notably Ireland and the European Data Protection Board) drive which of these practices become mandatory.
AI-written summary · about
chapters (11)
- 0:00Introduction
- 3:55Regulators' privacy focus
- 6:05Unusual names in data models
- 9:14Never delete data systems
- 15:16Necessary business data
- 17:44Algorithm logic in court
- 19:33Documentation for small biz
- 22:00Black box decision-making
- 24:08Pseudonymization rules
- 26:02GDPR risks by big tech
- 28:55Closing
related talks
01h 11mBack to Basics How Data Protection Principles Apply to AI Systems
01h 13mThe risk-based approach to privacy: How far should we go?
01h 14mSenior Academic Session I
29mKeynote Presentation: Minding the (Privacy Research/Practice) Gap
01h 02mThe GDPR and New Technologies: Enough, or do we need more?
37mQ&A with Keynote Katrina Riehl
01h 14mSenior Academic Session I.
13m"We are not Future-ready": Understanding AI Privacy Risks and Existing Mitigation...