The results of electronic elections should be verifiable so that any cheating is detected. To support this, many protocols employ an electronic bulletin board (BB) for publishing data that can be read by participants and used for verifiability checks. We demonstrate that the BB is itself a security-critical component that has often been treated too casually in previous designs and analyses. In particular, we present novel attacks on the e-voting protocols Belenios, Civitas, and Helios that violate some of their central security claims under realistic system assumptions. These attacks were outside the scope of prior security analyses as their verifiability notions assume an idealized BB.To enable the analysis of protocols under realistic assumptions about the BB, we introduce a new verifiability definition applicable to arbitrary BBs. We identify a requirement, called final-agreement, and formally prove that it is sufficient and, in most cases, necessary to achieve verifiability. We then propose a BB protocol that satisfies final-agreement under weak, realistic trust assumptions and provide a machine-checked proof thereof. Our protocol can replace existing BBs, enabling verifiability under much weaker trust assumptions.

In voting, disputes arise when a voter claims that the voting authority is dishonest and did not correctly process his ballot while the authority claims to have followed the protocol. A dispute can be resolved if any third party can unambiguously determine who is right. We systematically characterize all relevant disputes for a generic, practically relevant, class of voting protocols. Based on our characterization, we propose a new definition of dispute resolution for voting that accounts for the possibility that both voters and the voting authority can make false claims and that voters may abstain from voting.A central aspect of our work is timeliness: a voter should possess the evidence required to resolve disputes no later than the election's end. We characterize what assumptions are necessary and sufficient for timeliness in terms of a communication topology for our voting protocol class. We formalize the dispute resolution properties and communication topologies symbolically. This provides the basis for verification of dispute resolution for a broad class of protocols. To demonstrate the utility of our model, we analyze a mixnet-based voting protocol and prove that it satisfies dispute resolution as well as verifiability and receipt-freeness. To prove our claims, we combine machine-checked proofs with traditional pen-and-paper proofs.

In random sample voting, only a randomly chosen subset of all eligible voters are selected to vote. This poses new security challenges for the voting protocol used. In particular, one must ensure that the chosen voters were randomly selected while preserving their anonymity. Moreover, the small number of selected voters leaves little room for error and only a few manipulations of the votes may significantly change the outcome. We propose Alethea, the first random sample voting protocol that satisfies end-to-end verifiability and receipt-freeness. Our protocol makes explicit the distinction between human voters and their devices. This allows for more fine-grained statements about the required capabilities and trust assumptions of each agent than is possible in previous work. We define new security properties related to the randomness and anonymity of the sample group and the probability of undetected manipulations. We prove correctness of the protocol and its properties both using traditional paper and pen proofs and with tool support.

Many security protocols involve humans, not machines, as endpoints. The differences are critical: humans are not only computationally weaker than machines, they are naive, careless, and gullible. In this paper, we provide a model for formalizing and reasoning about these inherent human limitations and their consequences. Specifically, we formalize models of fallible humans in security protocols as multiset rewrite theories. We show how the Tamarin tool can then be used to automatically analyze security protocols involving human errors. We provide case studies of authentication protocols that show how different protocol constructions and features differ in their effectiveness with respect to different kinds of fallible humans. This provides a starting point for a fine-grained classification of security protocols from a usable-security perspective.