A Practical Implementation of Secure Auctions Based on Multiparty Integer Computation ⋆ Peter Bogetoft 1 , Ivan Damg˚ ard 2 , Thomas Jakobsen 2 , Kurt Nielsen 1 , Jakob Pagter 2 , and Tomas Toft 2 1 Department of Economics, Agricultural University, Copenhagen 2 Department of Computer Science, University of Aarhus Abstract. In this paper we consider the problem of constructing se- cure auctions based on techniques from modern cryptography. We com- bine knowledge from economics, threshold cryptography and security engineering to implement secure auctions for practical real-world problems. 1 Introduction The area of secure auctions combines three different areas of research: economics (mechanism design), cryptology, and security engineering. From economy and game theory, we know that many forms of auctions and trading mechanisms rely on/can benefit from a trusted third party (TTP), also known as a mediator or social planner. However, in a real application, it will often be the case that such a TTP cannot be found, or is very expensive to establish (since one basically has to counter-bribe it). Multiparty computation can be used to “implement” such a TTP in such a way that we only need to trust some fraction, say a majority, of the parties. Our goal is to investigate if this can also work in practice, and our work indicates that the answer is yes. In this paper we give an overview of practical cryptographic protocols which securely implements basic integer operations. Detail of these protocols can be found in [7] and [20]. We also give an overview of specific types of auctions which are practically realizable based upon these protocols. Detail of these auctions can be found in [2], but the details of the applications areas are held confidential due to commercial interests of the industry partners. Finally, we give a report on the empirical results from our prototype implementation. 2 Secure Auctions Secure auctions are emerging as a field of research in its own right. In recent years a number of contributions have been made (e.g. [10, 17, 3, 4, 21, 15]). ⋆ This work is sponsored by the Danish Research Agency. G. Di Crescenzo and A. Rubin (Eds.): FC 2006, LNCS 4107, pp. 142–147, 2006. c  IFCA/Springer-Verlag Berlin Heidelberg 2006