Privacy & Data Protection Ethiek en de rol van de Overheid

Transcription

1 Privacy & Data Protection Ethiek en de rol van de Overheid Jeroen van den Hoven Delft University of Technology Permanent Member European Group on Ethics, European Commission, Brussels Member High Level Ethics Group EDPS, Brussels

2 New Technology Ethics AI Deep Learning Robotics Big Data Quantum Computing Cloud Computing Block chain Internet of Things Social media and On-line platforms 3D Printing 5G Privacy Accountability Democracy Property Control Safety Security Sustainability Human dignity Identity Social cohesion

3 Verwarring Cyber Net E- Artificial Virtual Internet Digital Online Blockchain Community Friendship Person Intelligence Property Crime Democracy Education Organization

4 Vacuum Cyber community : a sort of community Digital Democracy : a sort of democracy Blockchain organisation : a sort of organisation Conceptual Vacuum Policy Vacuum Design Vacuum

5 1. Big Data & AI

6 Big Data

7 Convergence digital technologies Cloud Computing Social Networking Mobile Computing Robotics Augmented Reality Sensors Internet of Things

8 shopping Big data Social relations Preferences, emotions Location, movement

9 Twitter map

10 AI, Machine Learning, Neural nets Deep Learning

11 Algorithms used to sift through Big Data; find remarkable patterns Classify, categorize, screen, select, label, characterize, predict, decide, recognize, sense, distinguish, represent, choose, target, judge, sentence, refuse, nudge, coach, profile, select,

12

13

14 Algorithmic citizenship IUS Algorithmi

15 Aspire Health, USA Cost reduction in palliative care

16 2. High Stakes

17 Eric Schmidt CEO Google We know who you are, where you have been, more or less what you think

18 300 Likes = as good as your partner

19 Mind Reading

20

21

22 It is all about connecting People! Really?

23 Zuboff: Surveillance Capitalism

24 Big Data - Big Gov: Singapore

25

26 Citizen Scores

27 The Silicon Valley Model

28 The Russian Model

29

30

31

32 Macron on Responsible AI

33 The Good Big Data Society? Equality Liberty Solidarity Property Transparency Accountability Democracy Justice Privacy Dignity

34 Smart Surveillance City

35

36

37 Paul Nemitz Concentration of power with Big Tech; Everyone is so interested in ethics these days,..as long as it is not law.

38 Concentration of Power 1. Capital accumulation 2. Ownership of ubiquitous and central infrastructure/platforms 3. Domination of democracy and discourse 4. Control over persons and their data 5. Monopolies in AI science, technology and innovation

39 Nemitz: In order to protect and strengthen Western Liberal democracies in the Age of AI and the core trinitarian idea of human rights, rule of law and democracy we need a new culture of technology and business development which we call human rights, rule of law and democracy by design

40 3. Ethical Problems

41 Ethical Problems 1. Knowledge 2. Equality 3. Democracy 4. Social Justice 5. Privacy

42 Knowledge: Explainability, Transparency?

43 Machine Learning: Black Box

44

45 Fairness, Accountability, Transparency THE BLACK BOX SOCIETY

46 Democracy: Freedom, Autonomy.

47 Cambridge Analytica: Nix

48 Big Data Machinery

49

50 IBM: 47 psychographic parameters Includes values and needs

51 Polarization

52 Microtargetting

53 Big Data Big Nudging CHOICE ARCHITECTURES CASS SUNSTEIN

54 nudges

55

56 The brain is a 50 watt device that tries to save energy

57 A/B testing

58 Advanced Behavioral Science + Big Data + AI

59 Pre-programmed

60 Digital Society: having Control or being Controled

61 Advanced Behavioral Science + Big Data + AI

62

63

64 DATA PROTECTION: Met data kan je mensen Schade toebrengen Exploiteren Discrimineren Manipuleren Stigmatizeren Keuzen, autonomie en vrijheid beperken hun waardigheid aantasten

65 EU Data Protection Supervisor: Ethics Advisory Group Beyond GDPR

66 The law is not enough (EDPS Butarelli) DPIA DPO Codes of conduct Certification Reputation Risk Assessment CSR Issue management & Public Affairs Ethics: think and judge for yourself.what are your MORAL responsibilities?

67 The Mother of all Privacy Conferences

68 IDPCC October 2018: GDPR & Ethics 1 Fairness and human rights 2. Responsibility: accountability, awareness, vigilance 3. System Transparency and Intelligibility 4. Ethics by Design 5. Empowerment of each individual 6. Unlawful and illegitimate bias and discrimination should be prevented

69 EDPS Ethics Group: Digital Ethics Ethics by Design (e.g. Privacy by Design) Responsible Innovation

70 4. Ethics by Design & Responsible Innovation

71 Hostile architecture

72 Churchill

73 Values Built into information Systems Interfaces Infrastructures Algorithms Reputation Filter Reach Classification Categorization Ontologies Code Protocols Integrity constraints Identity Management Systems Authorization Matrix Procedures Regulations Incentive structures Auction mechanisms Voting mechanism Monitoring and inspection Architectures Governance arrangements

74 Responsibility Privacy Accountability Agency Autonomy Sustainability Safety Security Values Norms Laws Ideals Ethics Principles Express Implement Justify Audit Artefacts Architectures Standards Encryption Systems Infrastructure Algorithms Computers Oiltankers Airplanes Reactors Roads Internet Electricity Grids Hospitals Key Problem 21st Century: Value Sensitive Design

75 Design for X Design for privacy Design for security Design for inclusion Design for sustainability Design for democracy Design for safety Design for transparency Design for accountability Design for responsibility

76 Values hierarchy Values High Level supra-functional requirements Norms Policies Mechanisms Protocols Design requirements

77 Values Priva cyy Norms Risk mitigat ion Accou ntabilit y Data quality Securit y Design requirements Coarse grainin g Data clusteri ng Pseud onymi zation

78 If we do not design for our values.. Others will do it for us! Our words lack meaning We miss opportunities to Innovate

79 Privacy Enhancing Technology

80 9 Licht op de digitale schaduw woensdag 5 oktober 2016

81 9 Licht op de digitale schaduw woensdag 5 oktober 2016

82 Verantwoord Innoveren: EN & EN & EN & EN Als je morele verplichting hebt om A te doen (tellen) en je hebt morele plicht om B te doen (privacy te respecteren) dan heb je de morele plicht om er voor te zorgen dat je zowel A als B ook kunt doen (te innoveren).

83 Coarse graining through clustering Monreale et al. clustering spatio-temporal trajectories: so that it remains possible to gauge the traffic in a city (utility), while making it impossible to e.g. stalk an individual citizen (privacy).

84 K-anonymity

85 Differential Privacy

86 Discrimination detection supplementing privacy by design Algorithmic discrimination. Ruggieri & Pedreschi (2009, 2010) formal tools that detect implicit or explicit forms of discrimination in datasets (e.g.credit scores)

87 ELIZABETH DENHAM UK INFORMATION COMMISSIONER "I do not believe data protection law is standing in the way of your success" "It's not privacy OR innovation it's privacy AND innovation" (Debuutrede 29 juli, 2016)

88 Let s Innovate Responsibly!