Fair Secure Computation (or how can I gain strategic advantage by - PowerPoint PPT Presentation

Rump Session 2016 Fair Secure Computation (or how can I gain strategic advantage by breaking fairness) Alptekin Küpçü Koç University

Fair Secure Computation ALPTEKİN KÜPÇÜ Assistant Professor of Computer Science and Engineering

Secure Multi-Party Computation X 1 X 2 X 3 X 4 X 5 X 6 12 May 2016 Alptekin Küpçü - Koç University 2

Secure Multi-Party Computation X 1 X 2 X 3 X 4 X 5 X 6 12 May 2016 Alptekin Küpçü - Koç University 3

Secure Multi-Party Computation Y 1 Y 2 Y 3 Y 4 Y 5 Y 6 12 May 2016 Alptekin Küpçü - Koç University 4

Secure Multi-Party Computation Y 3 12 May 2016 Alptekin Küpçü - Koç University 5

SMPC in a Corporate Setting I OWN Y 3 YOU 12 May 2016 Alptekin Küpçü - Koç University 6

Ideal World X 1 X 2 X 3 X 4 X 5 X 6 12 May 2016 Alptekin Küpçü - Koç University 7

Ideal World Y 1 Y 2 Y 3 Y 4 Y 5 Y 6 12 May 2016 Alptekin Küpçü - Koç University 8

Real World X 1 X 2 X 3 X 4 X 5 X 6 12 May 2016 Alptekin Küpçü - Koç University 9

Simulator X 1 X 2 X 3 X 4 X 5 X 6 12 May 2016 Alptekin Küpçü - Koç University 10

Fairness Impossible in General  Assume a trusted Arbiter is available  Only trusted for fairness, not security  May collude with players  Should not learn input/output  Optimistically employed  Must be efficient (otherwise bottleneck) 12 May 2016 Alptekin Küpçü - Koç University 11

Fairness Impossible in General  Assume a trusted Arbiter is available  Only trusted for fairness, not security  May collude with players  Should not learn input/output  Optimistically employed  Must be efficient (otherwise bottleneck)  Ideal TTP Real Arbiter 12 May 2016 Alptekin Küpçü - Koç University 12

Fair and Secure Computation  Fairness extentions and Arbiter resolutions must be simulated 12 May 2016 Alptekin Küpçü - Koç University 13

Simulating Fairness X 1 X 2 SECURE 2PC SIMULATION FAIRNESS ARGUMENT 12 May 2016 Alptekin Küpçü - Koç University 14

Simulating Fairness X 1 X 2 SECURE and FAIR 2PC SIMULATION 12 May 2016 Alptekin Küpçü - Koç University 15

Fair and Secure Computation  Fairness extentions and Arbiter resolutions must be simulated  Otherwise the protocol may be insecure! 12 May 2016 Alptekin Küpçü - Koç University 16

Fair and Secure Computation  Fairness extentions and Arbiter resolutions must be simulated  Otherwise the protocol may be insecure!  Simulator may contact only when fairness is guaranteed 12 May 2016 Alptekin Küpçü - Koç University 17

Simulator X 1 X 2 X 3 X 4 X 5 X 6 12 May 2016 Alptekin Küpçü - Koç University 18

Ideal World Y 1 Y 2 Y 3 Y 4 Y 5 Y 6 12 May 2016 Alptekin Küpçü - Koç University 19

Real World 12 May 2016 Alptekin Küpçü - Koç University 20

Fair and Secure Computation  Fairness extentions and Arbiter resolutions must be simulated  Otherwise the protocol may be insecure!  Simulator may contact only when fairness is guaranteed  Otherwise real and ideal world outputs are distinguishable 12 May 2016 Alptekin Küpçü - Koç University 21

Fair and Secure Computation  Fairness extentions and Arbiter resolutions must be simulated  Otherwise the protocol may be insecure!  Simulator may contact only when fairness is guaranteed  Otherwise real and ideal world outputs are distinguishable  Arbiter cannot harm security 12 May 2016 Alptekin Küpçü - Koç University 22

Our Solutions # Participants # Rounds # Messages 2 O(1) O(1) n O(1) O(n^2)  OPTIMAL asymptotic performance  Cut-and-choose or zero-knowledge  Malicious or covert  2PC or MPC 12 May 2016 Alptekin Küpçü - Koç University 23

Comparison  Compared to related works, we provide  Optimal asymptotic performance  Constant round (not gradual release)  No broadcast  Arbiter load independent of the circuit size  Do not require an external payment mechanism  In a competitive corporate setting, how can one value some output that is unknown beforehand?  Full simulation proofs  Arbiter cannot harm security  Also proven via simulation  Only fairness is lost if Arbiter colludes with malicious parties 12 May 2016 Alptekin Küpçü - Koç University 24

Our Papers  Reading  Kılınç and Küpçü, CT -RSA 2015, Optimally Efficient Multi-Party Fair Exchange and Fair Secure Multi-Party Computation  Kılınç and Küpçü, FC 2016, Efficiently Making Secure Two-Party Computation Fair  Küpçü and Mohassel, FC 2016, Fast Optimistically Fair Cut-and-Choose 2PC  Funding Acknowlegements  TÜBİTAK, the Scientific and Technological Research Council of TURKEY  COST Action IC1306 Cryptoaction 12 May 2016 Alptekin Küpçü - Koç University 25

ALPTEKİN KÜPÇÜ Assistant Professor of Computer Science and Engineering http://crypto.ku.edu.tr