Open Universiteit

Transcript

1. Privacy by design Challenges and Opportunities using Privacy Enhancing Technology (PET) Yanick Dols Ecosystem Developer at Brightlands Fabian van den Broek Assistant Professor at Open University
2. Yanick Dols Ecosystem Developer at Brightlands Smart Services Campus
3. Brightlands A Limburg initiative for a sustainable and healthy future Stimulate innovation Create employment Attracting and retaining talent Boosting the Limburg economy
4. From four innovation campuses we look for solutions for tomorrow that already lead to opportunities today Sittard-Geleen Brightlands Chemelot Campus Smart materials Sustainable production of chemicals Maastricht Brightlands Health Campus Regenerative medicine Precision medicine Innovative diagnostics Venlo Brightlands Greenport Campus Food Healthy nutrition Heerlen Brightlands Smart Services Campus Data science Smart services
5. Brightlands Chemelot Campus Brightlands Health Campus Brightlands Greenport Campus Brightlands Smart Services Campus
6. Knowledge crossing borders Geographically, organizationally and scientifically
8. We encourage close cooperation between entrepreneurship, science and governments Entrepreneurs hip Public sector Knowledge institutes Brightlands
9. We combine on our campus science, entrepreneurship, talent and state-of-the-art facilities Offices & Facilities Innovation & Research Talent & Education Public sector Citizens Entrepreneur s Knowledge institutes
11. You learn and innovate on a specific theme with our business and knowledge partners ESG reporting Poverty & debt Admin tech Smart public services Energy transition Climate change Fraud detection Digital identity HR tech Sustainable homes Digital inclusion Customer interaction Businesspartners Themes Knowledge Partners
12. WE WORK WITH SEVERAL Artificial Intelligence Synthetic data Multi Party Computing DISRUPTIVE TECHNOLOGIES Blockchain Self Sovereign Identity Quantum Computing
13. Fabian van den Broek Assistant Professor at Open University
14. security privacy the Security vs. Privacy debate security privacy
15. security privacy
16. security problems privacy problems
17. security problems privacy problems 1 1. e-mail scam
18. security problems privacy problems 1. e-mail scam 2. account take-over by criminal 2
19. security problems privacy problems 1. e-mail scam 2. account take-over by criminal 3. Google employee spying on your e-mail 3
20. security problems privacy problems 1. e-mail scam 2. account take-over by criminal 3. Google employee spying on your e-mail 4. profiling by Google based on your e-mail 4
21. security problems for Google privacy problems for users 1. e-mail scam 2. account take-over by criminal 3. Google employee spying on your e-mail 4. profiling by Google based on your e-mail security problems for users
22. security problems for Google privacy problems for users 1. e-mail scam 2. account take-over by criminal 3. Google employee spying on your e-mail 4. profiling by Google based on your e-mail 5. cryptominer on Google’s servers security problems for users 5
23. 1. e-mail scam 2. account take-over by criminal 3. Google employee spying on your e-mail 4. profiling by Google based on your e-mail 5. cryptominer on Google’s servers security problems privacy problems
24. 1. e-mail scam 2. account take-over by criminal 3. Google employee spying on your e-mail 4. profiling by Google based on your e-mail 5. cryptominer on Google’s servers different attacks & different attackers
25. 1. e-mail scam 2. account take-over by criminal 3. Google employee spying on your e-mail 4. profiling by Google based on your e-mail 5. cryptominer on Google’s servers different goals
26. 1. e-mail scam 🡪 finances 2. account take-over by criminal 🡪 authenticity/integrity 3. Google employee spying on your e-mail 🡪 privacy / confidentiality 4. profiling by Google based on your e-mail 🡪 privacy / autonomy 5. cryptominer on Google’s servers 🡪 computation cycles
27. 1. e-mail scam 2. account take-over by criminal 3. Google employee spying on your e-mail 4. profiling by Google based on your e-mail 5. cryptominer on Google’s servers How to compare or value these?
28. security privacy
29. less security more security
30. Helps keep data confidential Helps availability of data backups
31. security privacy
32. good for security property Y that party A & B care about assuming attacker model E bad for security property Y’ that A & C care about assuming attacker model E’ good for security property X of party A assuming attacker model E’ bad for privacy property X of party B good for privacy property Z’ of party B bad for societal concern Z’’ design space
33. security requirement Y detecting abuse usability • for users & customers • for organisation • for sys-admins • for software developers cost security requirement X revocation non-repudiation anonymity privacy requirement Z repudiation benefits
34. Do not despair…. • Security is hard… • Privacy is (often) harder… • But, you have to do it. By design.
35. Enter PETs….
36. Privacy Enhancing Technologies (PETs) Techniques that improve privacy, while keeping functionality
37. Many different PETs • Zero knowledge proofs (ZKP) • Attribute-based Credentials (ABC) • Fully Homomorphic Encryption (FHE) • Multi-Party Computation (MPC) • Federated Learning (FL) • Differential Privacy (DP) • ….
38. Many different PETs • Zero knowledge proofs (ZKP) • Attribute-based Credentials (ABC) • Fully Homomorphic Encryption (FHE) • Multi-Party Computation (MPC) • Federated Learning (FL) • Differential Privacy (DP) • ….
39. Zero-Knowledge Proofs (ZKP) Proof knowledge of a certain value, without revealing it • Pick a card from a randomized deck • Prove the color of the card (red/black), without revealing anything else
40. Zero-Knowledge Proofs (ZKP) Prover Verifier Proof Prover secret ?
41. ZKP
42. ZKP ABC Attribute-based credentials (ABC) • Proof ‘attributes’ of yourself • Name, birthdate, bloodtype, etc. • Without revealing anything else
43. ZKP ABC Attribute-based credentials (ABC) • Proof ‘attributes’ of yourself • Name, birthdate, bloodtype, etc. • Without revealing anything else
44. Multi-Party Computation • Allow multiple parties to compute something together • Each party has their own input • No party learns the input of other parties • The result can be shared, or individualized
45. … 1 secret 2 3 n secret secret secret function
46. So many PETs…
47. PET Decision Tree
51. More information? Step by at our stand #31 Thank you!