Is Arizona's Internet Voting System Safe Enough? 171
JMcCloy writes "Kevin Poulsen, senior editor at Wired News, asks readers 'Is internet voting safe?' and has a poll at the end of the article. So far, 32% responding actually think that internet voting is worth it, risks and all. It is scary how easily people can be persuaded to trust a system that is so vulnerable." The system described, used in Arizona in last year's election process, isn't just checking a box and clicking a button, but Poulsen lays out some scenarios by which it could be subverted.
Scary? (Score:2, Informative)
So far, 32% responding actually think that internet voting is worth it, risks and all. It is scary how easily people can be persuaded to trust a system that is so vulnerable."
So you're saying that an internet poll (something that's guaranteed to have a bias towards everything internet) has a strong majority of people agreeing that internet voting is not worth it, and the conclusion you reach is that "[it's] scary how easily people can be persuaded to trust a system that is so vulnerable?" The numbers seem to suggest that it actually isn't all that easy to persuade people to trust such a system.
Not much different than mail in ballots (Score:5, Informative)
Whereas "true" Internet voting is a phenomenally bad idea (when implemented in a way that's acceptable to the majority of voters), the Arizona system isn't really Internet voting. It's more "absentee ballots" that use the Internet as the delivery mechanism rather than the normal postal system.
Mail-in ballots are extremely common in Arizona ever since they changed the "absentee balloting" system into a more generic "everybody can use it" system. For instance, I have a ballot automatically mailed to be before every election, no matter how big or small, without me having to do anything but sign up a couple years ago. It's very slick.
The ballot is a normal paper one exactly like those found in the polling place. I fill it out by completing arrows pointing to my choice (easy and not even remotely ambiguous) then put it in a specially coded envelope that I sign and mail in. On the other end, a poll worker opens the envelope, marks that I voted (to prevent multiple votes), saves off my signature, and puts the ballot through the normal recording devices to record my vote. The voter lists in my local polling place have me marked as "mail in" so if I were to drop by on election day, they would accept my ballot but it would only be counted after all other ballots are counted and they can verify that I hadn't already voted.
It's extremely convenient and has made the difference between voting only in the major elections to voting in all of them (and learning a lot more about local candidates in the process). The drawback is that I have to trust that my vote isn't tied with my name. See, when you are at a polling station, then they record that you voted, but your actual ballot isn't in any way tied to you. With the mail-in process, it's possible that that is still the case (maybe the person/system opening the envelopes isn't the one recording the votes)... but you can't know for sure. For all I know, they may have a database mapping people with who they vote for. Honestly, that doesn't bother me at this point. I am pretty vocal about who I vote for and have even publicly posting my voting lists for the world to see before. I guess I would stop the mail-in only if I had reason to believe that my vote wasn't being counted.
Anyway, that's the mail-in system. The "Internet voting" system is effectively that but for people overseas. That option was never available for me since I'm local. The only difference is that instead of putting their ballot into an envelope and signing that, they instead scan it in and upload it to a server. Everything else is identical.
The article does make a few good points on some ways that that system could be subverted. Yeah, there are definitely a few more attack points... but they seem a little far fetched at this point. The level of effort required to implement any of the attack vectors would only be worth it if done at a bigger scale. That is, if this started being available to ALL AZ residents, then it starts to matter. For now... meh.