I recently signed up for two different mobile banks (Monzo and bunq). Both of them verified my identity by having me take a photo of my passport, but then also recording a video of myself (with my face clearly visible) saying an exact phrase they specified, in one case “My name is $legalname and I'd like to open a Monzo account” and in the other a sequence of random numbers.
That seems way more secure.
The problem with voting is not and has never been physical people people voting fraudulently. It is the alteration of the votes on a central machine. (For example as an easily editable excel file on at least on occasion)
The extra barriers to confirm the physical person standing at the booth is a redirection of a serious problem with Putin into a little more voter suppression.
Face2Face would defeat this: anyone with enough images of the victim (and copy of ID) could impersonate them in this scheme.
It's kinda' not, because of practicality. There's enormous risk and time involved to get that one erroneous vote. Who's going to do that for a vote that has a smaller chance of swinging an election than you have of winning the lottery?
If you have a stray actor doing that on election day they maybe vote 10 times if they're a logistical genius doing an Ocean's 11 of voter fraud. That's enormous risk of actual jail for still almost zero chance of actually swaying an election. That's why no one does that.
And the idea that you're going to coordinate that effort? You've increased the odds of possibly being statistically significant a tiny iota but have exponentially increased the chances of being caught.
This is why voter fraud arguments are complete bullshit and just cover for voter suppression.
Not speaking to anything else related to this at all, it does sound like they are quite a fly by night operation based on everything that is being dug up.
However, the technology mentioned above is not something they came up with themselves, and in fact see wide spread use for purposes of fraud prevention for financial systems (which sees active attacks against verification tools). Its not as easy to fool as you are making it seem.
This tweet chain is pretty interesting:
I'm just speaking about the technology as described. Its not a novel id verification method, plenty of other companies use it and there are multiple vendors / providers of the service.
Different organizations have different security needs
Financial systems can use the relative cost of a security measure compared to the cost of a breach to decide whether to implement the security; they consciously accept a certain level of failures. Also, they can use insurance to mitigate exploits after they happen.
Voting systems cannot tolerate failures, at least in theory. Every vote must count. Also, at least for some elections, the value to the attacker is much higher than what can be gained from financial institutions; a foreign intelligence service has much more to gain than money from controlling a US presidential election, for example, and an investor could make billions by knowing the outcome ahead of time. Finally, there is no mitigation after a breach: Insurance won't save a fraudulent election.
The point I am making is that its not as simple as described to "spoof" this tool, and to give some background on its usage since it seemed to be a novel verification tool for the OP
> every vote can potentially be crucial
I'll also add that every vote is important because the number of votes is important. Winning 70-30 will get much different behavior than winning 50.01-49.99; the latter representative had better pay attention to the 49.99's needs. And that can matter down to the smallest geographical levels - if everyone in your neighborhood votes, politicians will pay much more attention.
So it sounds like they verify your identity based on two photos that you provide? So if two photos provided by an unknown person match, then they become a trusted person? Even if there's a separate step that matches that to the DMV database or something... all you actually need is a photo of the person you want to impersonate.
But pretending to be someone else for remote voting is already a weak point, and one that would be expensive to exploit on a large enough scale to make a difference. The much much bigger risk is that we have absolutely no way to verify that the votes recorded are correct. This company's app sends the vote to the company's server which stores it in this company's database. That's three steps at which the votes can be easily changed.
The question to ask with any proposed voting system is: how can we verify that the counts are accurate? The _only_ way to do that in anything close to a trustworthy manner, is by having an established network of trusted agents (one or more witnesses from each interested party organization at each physical voting location) monitor a human-visible process for collecting paper ballots which can be counted and recounted at will by multiple groups of interested parties.
Any system where the votes are ever hidden away from witnesses and accessible, say, in a back room with an unlocked door leading to an empty alleyway ... or on any computer system, is inherently insecure.