Blockchain Voting: Solves None Of The Actual Problems Of Online Voting; Leverages None Of The Benefits Of Blockchain


Just recently we wrote about why blockchain-based DRM was a terrible idea, and it could be summed up by the simple fact that a blockchain solves none of the “problems” of DRM today, and leverages none of the actual benefits of a blockchain. And… now I feel like writing basically the same exact post around blockchain voting. Like blockchain DRM, blockchain voting is one of those ideas that gets tossed around a lot. For decades, lots of people who actually understand computer security have explained why online voting is a horrifically bad idea in that it involves effectively unsolvable problems. It’s not that it’s a “hard” problem, it means that online voting is effectively impossible without massive changes to almost everything we do in ways that we can’t really comprehend right now. There are some serious researchers who are thinking about this, but to date, there is nothing even remotely close to to being acceptable, and there may never be.

And yet, the “simplest” way that some people understand the risks of online voting is basically “it would be bad if someone could change your vote and no one would know.” That’s an easy to understand point to make, but the problems with online voting go way, way beyond that. Do a simple Google search on why online voting is a terrible idea and you’ll get dozens of on-point results, but if you want a nice, simple explanation of just the first pass of potential risks with online voting, check out this video from a couple years ago by Princeton professor Andrew Appel, who has been studying voting security for many, many years:

It’s 21 minutes, and if you’re unsure of why internet voting is dangerous or think there’s a simple solution, I’d urge you to watch it. But for those who don’t, I’ll just toss up one single slide from the presentation, which is not even remotely comprehensive in the list of potential problems with online voting:

That doesn’t even get at a number of other potential issues (some of which are discussed in the video). And yet — as with blockchain-for-DRM — there’s always someone who thinks that the only real problem is the double spend problem. Enter Alex Tapscott and the NY Times. Alex Tapscott is the son of Don Tapscott, who has written a number of fairly influential books related to technology and innovation, including “Growing up Digital” and “Wikinomics.” In 2016, he teamed up with his son, Alex, and wrote a book called “The Blockchain Revolution,” which is a fun read (they sent me a copy), if a bit overly excited in its analysis of potential implementations of the blockchain. As I’ve said in the past, I’m a believer that blockchain/tokens can completely revolutionize a few areas of the internet, but people have yet to really figure out which areas can take advantage of what is unique about the blockchain (beyond highly volatile currencies).

My favorite review of the book on its Amazon page includes this lovely sentence: “After the opening chapter, it turns into a rambling acid trip of delusional fantasies about exactly how blockchain will inevitably fix all the things wrong with society and the world.”

Anyway, along comes Alex Tapscott and on election day, the NY Times gave him precious space to spew utter nonsense about how it’s time for online voting… via the blockchain.

The key weakness of early online voting systems was the inability to solve what cryptographers called the “double spend problem.” When we send a file on the internet, we’re actually sending a copy of that file; the original remains in our possession. This is acceptable for sharing information but unacceptable for recording votes in elections. The possibility that individuals could cast their ballots multiple times for a candidate made these systems useless — just as vulnerable as paper ballot systems. Points of failure included susceptibility to hackers, coding bugs, and human error. With enough resources, any rogue could “stuff” a digital ballot box with illegitimate votes.

Except… that’s not the key weakness in early online voting systems. It is one problem, but kinda far down the list. Look at that still from Appel’s video above. Double spending isn’t even there, really. Yet, Tapscott’s piece acts as if it’s the biggest problem, and easily solved with blockchain.

Since the NY Times published that article, plenty of folks with actual computer security expertise have stepped up to debunk it. Ben Adida, the Executive Director of a new organization called Voting Works, attempting to build secure, open source voting machines, actually debunked it a year ago (that’s how good he is):

In a typical election setting with secret ballots, we need:

  1. enforced secrecy: a way for each voter to cast a ballot secretly and no way to prove how they voted (lest they be unduly influenced)
  2. individual verifiability: a way for each voter to gain confidence that their own vote was correctly recorded and counted.
  3. global verifiability: a way for everyone to gain confidence that all votes were correctly counted and that only eligible voters cast a ballot.

Let’s say we have a Blockchain-style distributed database. How far does that get us to meeting these needs?

A distributed database of all cast votes, where everyone sees the same state of the world, would certainly be useful for (3) global verifiability and to some degree for (2) personal verifiability. That said, it won’t get us all the way there on those, and it won’t get us anywhere on (1) enforced secrecy.

Specifically, to combine personal verifiability with enforced secrecy, we need some mechanism that gives each voter enough confidence that their vote made it all the way to the tally, but not so much that they can sell their vote to a buyer/coercer. A public ledger of plain votes is a terrible idea, since that makes vote selling trivial. A public ledger of vote tracking numbers of sorts is better for privacy, though it doesn’t really provide actual verifiability that the contents of the ballot weren’t tampered with. Clearly, we need something more, and that something simply isn’t provided by a distributed ledger.

In a typical election setting with secret ballots, we need:

  1. enforced secrecy: a way for each voter to cast a ballot secretly and no way to prove how they voted (lest they be unduly influenced)
  2. individual verifiability: a way for each voter to gain confidence that their own vote was correctly recorded and counted.
  3. global verifiability: a way for everyone to gain confidence that all votes were correctly counted and that only eligible voters cast a ballot.

Let’s say we have a Blockchain-style distributed database. How far does that get us to meeting these needs?

A distributed database of all cast votes, where everyone sees the same state of the world, would certainly be useful for (3) global verifiability and to some degree for (2) personal verifiability. That said, it won’t get us all the way there on those, and it won’t get us anywhere on (1) enforced secrecy.

Specifically, to combine personal verifiability with enforced secrecy, we need some mechanism that gives each voter enough confidence that their vote made it all the way to the tally, but not so much that they can sell their vote to a buyer/coercer. A public ledger of plain votes is a terrible idea, since that makes vote selling trivial. A public ledger of vote tracking numbers of sorts is better for privacy, though it doesn’t really provide actual verifiability that the contents of the ballot weren’t tampered with. Clearly, we need something more, and that something simply isn’t provided by a distributed ledger.

That’s only part of Adida’s thorough takedown of the concept.

Tim Lee at Ars Technica highlighted another batch of problems:

Tapscott focuses on the idea that blockchain technology would allow people to vote anonymously while still being able to verify that their vote was included in the final total. Even assuming this is mathematically possible—and I think it probably is—this idea ignores the many, many ways that foreign governments could compromise an online vote without breaking the core cryptographic algorithms.

For example, foreign governments could hack into the computer systems that governments use to generate and distribute cryptographic credentials to voters. They could bribe election officials to supply them with copies of voters’ credentials. They could hack into the PCs or smartphones voters use to cast their votes. They could send voters phishing emails to trick them into revealing their voting credentials—or simply trick them into thinking they’ve cast a vote when they haven’t.

[…]

But let’s think about how this would play out in practice. Suppose it’s mid-November 2020 and Donald Trump has narrowly won reelection. A few thousand voters in key swing states come forward to say that they intended to vote for Trump’s opponent but their vote was recorded for Trump instead. Thousands of others say they tried to vote for Trump—or against him—but their votes weren’t counted.

Was that due to hackers meddling with the vote, technical snafus, or user error? Were some of them just misremembering how they had cast their ballots? There would be no way to know for sure.

An important property for an election is finality: you want a well-understood process that makes people confident in the result. The paper-based process used in most states today isn’t perfect, but it’s pretty good on this score. Each vote is recorded on a paper ballot that’s available for anyone to look at. Everyone understands how paper ballots work. People can observe the vote-counting process to verify that no ballots were altered. So not only does the process usually lead to an accurate count of peoples’ votes, it also builds public confidence in the integrity of the result.

Blockchain voting would be much, much worse. Hardly anyone understands how a blockchain works, and even experts don’t have a good way to observe the online voting process for irregularities the way an election observer does in a traditional paper election. A voter might be able to use her private key to verify how her vote was recorded after the fact. But if her vote wasn’t counted the way she expected (or wasn’t counted at all) she’d have no good way to prove that she tried to vote a different way.

Just a few months back, we also wrote about the terrible idea that West Virginia was experimenting with, via a company called Voatz (which is mentioned in Tapscott’s article) that was building a “blockchain-based” system to allow military personnel overseas to vote via their mobile phones. And of course, as we noted at the time, it had all the same problems of all these systems. What it adds in “convenience” (if anything) is completely outdone by the security nightmare it creates.

Again, I still think blockchains have some potential to do some pretty useful things, but the idea that they can solve any old basically impossible under current realities technology problem by sprinkling magic “crypto” and “distributed” pixie dust on the problem is not a good look. Which should lead people to asking why the NY Times is publishing it without any fact checking at all?



Source link

0 replies

Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply

Your email address will not be published. Required fields are marked *