During the last decade, Internet-voting (i-voting) moved from the field of fundamental research to practical application. First, we will see that theoretical research provides satisfying algorithms for some of the challenges raised by i-voting and that some real world experiments have already been developed and performed.
Unfortunately, in current i-voting systems, the citizen loses his/her control over the overall electoral process. Indeed, only insiders usually have access to the programming code of the application and to the servers used in i-voting. The confidence in democracy itself could be harmed by this opacity.
The European Convention on Human Rights emphasizes that votes should remain secret. This can not be assured for i-voting, since it is not possible to have a booth around each computer for example during the voting process. Family voting cannot be prevented and vote buying could be a major threat for democracy.
Moreover, we can not assume that the voter's computer contains no viruses or Trojan horses. Therefore, it is optimistic to assume that the ballot transfered to the server is the one chosen by the voter.
Finally, we will see that the effect of i-voting on the turnout at polls might remain marginal.