The book. For or against electronic voting? At each election, the question comes up, with its share of arguments: modernity, speed, fight against abstention… Thanks to two IT specialists, specialists in this subject, the voter will understand to what extent this question is badly posed, and that here as elsewhere technology will not be able to solve everything.
The list of problems to be resolved is indeed very long and the beginning of the book consists of setting out the definitions and exposing the constraints. Obviously, the secrecy of the vote is unavoidable. The usability and ergonomics of the device are also priorities.
On the other hand, more delicate are the questions of verifiability, that is to say the techniques implemented to control the vote. Checking that your ballot is in the ballot box is one thing, checking that the machine (often our personal computer) has indeed voted in accordance with our intention is much more complex. Similarly, we may want any outside person to be able to check the count. If there are technical solutions to these problems, they are not simple and the prerequisite should be transparency, ie the exact description of what the voting algorithm is doing. This is not always the case, as the authors explain in the chapters devoted to the law and audits of these systems.
The mysteries of computer security
These first chapters also bring in the complex mysteries of computer security researchers, who, without showing paranoia, are very imaginative in proposing twisted tricks or showing that something simple in appearance may not be so in practice. For example, being able to decrypt bulletins with a key seems convenient. But then you have to trust only one person. Having several keys corrects the defect, but then losing one makes it impossible to announce the results. So how many do you need?
It is also necessary to underline two excellent chapters which expose the various solutions resulting from cryptography. Not only are they very clear, but they also describe techniques used in other aspects of our digital life, which is always good to master. Symmetric, asymmetric, threshold, homomorphic encryption, signatures, mixers… are thus detailed. Just like the astonishing concept of zero-knowledge proof (“zero-knowledge evidence”). And too bad for the surprise: do not rely too much on the blockchain to ensure a reliable vote.
You have 19.83% of this article left to read. The following is for subscribers only.