ISRI Seminar Trustworthy Electronic Election Results without Trusted Machines C. Andrew Neff, Vothere.net Wednesday, November 17, 2004, 12 pm, Wean 4623 lunch will be provided Abstract We present a cryptographic protocol that achieves universal verifiability in elections; that is, each voter can determine from the final tally that her vote was properly counted. Furthermore, any member of the public can total the results herself. The voting machines used need not be trustworthy. Using a challenge-response scheme, the voting device is prevented from casting a ballot which is inconsistent with the voter's intent without showing evidence which the voter can easily detect by simple inspection. To prevent vote tampering after ballots have been cast, each voter is given a receipt which can be used to audit the public ballot box. However, because part of the voter's proof of ballot correctness is derived from direct observation in the poll booth of the temporal sequence by which the receipt is formed, the receipt is meaningless to someone else and thus cannot be used to prove to another party how the voter voted. Voters can track their own ballots through the final count and dispute any discrepancy. A commercial implementation of this system is offered by VoteHere (see www.votehere.net). Bio: Dr. Neff joined VoteHere in 1999. Previously he was a Research Staff Member at the IBM T.J. Watson Research Center. While there, he developed and published several new ideas in the areas of algebraic and geometric computation and Design of Experiments. In 1990 and 1996 he solved a fundamental problem determining the optimal complexity of solving non-linear univariate equations which had been an open problem, important to several fields of computer science, since the mid 1970s. He left IBM to work for Stratasys, a company developing a "3D printing" machine for direct forming of solid parts from electronic data. There he improved on state-of-the-art algorithms and software for geometric computations, and dynamic control. Recently, he has developed protocols for conducting secure, secret ballot, digital elections - most notably improved methods for verifiable mixing of encrypted data. Dr. Neff received his BS and MS from the University of Chicago, and his Ph.D. in theoretical mathematics from Princeton University.