[HN Gopher] Ask HN: What is your opinion on electronic voting?
       ___________________________________________________________________
        
       Ask HN: What is your opinion on electronic voting?
        
       Here in Brazil there is somewhat of a debate if electronic voting
       (for presidency, senate, everything) is a good thing. We have been
       using it since 1996 here, and most people are fine with it. There
       is no evidence of fraud since then, but I don't know how easy it
       would be to know for sure. I've seen some opposition to it, like
       Tom Scott's videos[1][2], but I also remember seeing knowledgeable
       people being in favor of it (though I can't remember where). I have
       a programming background, so personally I lean against it, but I
       don't know enough about security at this level to have a meaning
       opinion. I would like to know what everyone here thinks, especially
       those with a career on cyber-security.  P.S.: I'm not talking about
       online voting here, which is a whole other can of worms. I'm only
       referring to electronic machines used to vote, without any internet
       connection.  [1] https://www.youtube.com/watch?v=w3_0x6oaDmI  [2]
       https://www.youtube.com/watch?v=LkH2r-sNjQs
        
       Author : strcspn
       Score  : 7 points
       Date   : 2022-09-14 22:21 UTC (40 minutes ago)
        
       | Schroedingersat wrote:
       | Involving a bunch of people is a security feature. Perhaps the
       | only one that works consistently.
       | 
       | The only reason to move to electronic voting is to reduce the
       | number of people. Ie. an attempt to reduce the security of the
       | election.
       | 
       | Additionally your voting system has to he simple enough for
       | transparency to be effective. Ie. a non-compromised voting system
       | must be distinguishable from a compromised one by the average Joe
       | or it will get replaced by the compromised one.
       | 
       | No electronic voting system passes this test.
       | 
       | By all means use a computer to help (ie. optionally mark the
       | paper with a computer, or do some zero knowledge vote
       | verification thingy or count the marks automatically as a pre-
       | count), but it should never interfere with the basic process of
       | put mark on bit of paper, put bit of paper in box, have human
       | tally marks in box.
        
       | tarakat wrote:
       | Not even formally verified open-source compilers can be trusted
       | (and for the record, in practice, voting machines do not reach
       | anywhere near such level of security):
       | https://cs.cmu.edu/~rdriley/487/papers/Thompson_1984_Reflect...
       | 
       | Then you also have to trust the entire supply chain (software AND
       | hardware):
       | https://www.schneier.com/blog/archives/2018/03/adding_backdo...
       | 
       | Because the prize is so great, the system has to withstand
       | decades-long attacks by global superpower-level actors. And
       | unlike a bank account, you don't know when you've been hacked.
       | 
       | Once compromised, a smart attacker will stay within the public
       | poll's margin of error - which is much greater for small, local
       | elections through which most politicians have to pass. Think 30%
       | or more, in countries where people have little to fear from
       | answering such polls truthfully. Or venture outside of it,
       | deliberately making the election look stolen to throw a country
       | into chaos.
       | 
       | No system that relies on an un-auditable black box for security
       | can be trusted with elections. This is so blindingly obvious,
       | anyone that says otherwise, no matter how credentialed, is either
       | deluded, corrupt, or malicious.
       | 
       | All this risk, and the benefit is.. what? Countries all over the
       | world manage to vote and tally the results quickly using nothing
       | but paper.
       | 
       | > I'm only referring to electronic machines used to vote, without
       | any internet connection.
       | 
       | If you don't trust the chip fab, how do you check there is no
       | connection? I suspect antennas can be hidden incredibly well, and
       | there are countless side-channels that can be used for
       | communication. And even without antennas, machines are not safe:
       | _It is typically introduced to the target environment via an
       | infected USB flash drive, thus crossing any air gap._ -
       | https://en.wikipedia.org/wiki/Stuxnet
       | 
       | Do you trust every PC and every USB drive that comes into contact
       | with the voting machine? And every PC _that_ PC came into contact
       | with, and so on?
       | 
       | Edit: To clarify, paper isn't immune to tampering. But crucially,
       | tampering with it is much harder to hide, and it _doesn 't
       | scale_. Not unless you already have loyal agents controlling most
       | polling stations in a country - at which point, no voting system
       | can save you. But even then, _hiding_ your activity will be hard,
       | unlike if you only have to flip a few bits.
        
       | type0 wrote:
       | My opinion is simple: it shouldn't exist in democratic
       | institutions. It is fine if you vote among the members of your
       | hobby organization.
        
       | JohnFen wrote:
       | I think electronic voting is a disaster waiting to happen.
        
       | zxwrt wrote:
       | You can't hack the paper, but you can hack the machine, no matter
       | how secure it is...
        
         | Thorentis wrote:
         | Well, it's also much harder to crack down on voter fraud like
         | voting on behalf of dead people. I would say paper voting has a
         | larger attack surface _for the average voter_.
        
           | Schroedingersat wrote:
           | Those things are overwhelmingly less common than you are
           | pretending, make very little difference, are substantially
           | easier to prevent than attacks on electronic voting, and are
           | trivial to detect.
        
       | ksaj wrote:
       | I personally think the risks are less than for bank machines. But
       | only if lessons learned in the development and maintenance of
       | bank machines is fully embraced for the voting machines.
       | 
       | There are some banks that let you deposit cheques without needing
       | to type anything - they use OCR to find the amount, then verify
       | with you, then allow you to print a copy of the cheque with the
       | receipt. There is no reason this same technology can't be used in
       | voting machines, with the difference being that the visual proof
       | is kept internally instead of being printed out. That visual
       | proof is what is needed in case of a Trump-style fishing audit.
        
       ___________________________________________________________________
       (page generated 2022-09-14 23:02 UTC)