This afternoon on my Pacifica Radio show on KPFK here in Los Angeles (and up and down the West Coast) I interviewed Dr. Roger Johnston of Argonne National Lab’s Vulnerability Assessment Team on their $26 “Man-in-the-Middle” attack on a Diebold touch-screen voting system as I reported yesterday in my exclusive at Salon (and here at The BRAD BLOG.)
While yesterday’s reports included information on the team’s previous, similar-ish hack of a Sequoia touch-screen system, and the fact they feel they could do the same with most other e-voting systems, among the interesting new points raised during our discussion on KPFK today is that Johnston believes they could use a similar method to manipulate the vote count on a paper-ballot based optical-scan system as well!
So, in other words, the low-rent attack Argonne has demonstrated — requiring no knowledge of the voting system software, $10 to $26 in off-the-shelf computer parts, and little more than an 8th grade computer lab education — could also be implemented not just on touch-screen e-voting systems, with or without a so-called “paper trail,” but also on the paper ballot op-scanners used to count the majority of votes that will be cast in the U.S. in next year’s Presidential election.
(And if you’re now wondering: “Well, what the hell should we be using to tabulate our elections?!” I’ll refer you to this “Democracy’s Gold Standard” special comment of mine, as taken from a recent Mike Malloy Show which I posted a few months ago.)
Here’s my interview with Dr. Johnston today…
Download MP3 or listen online here [appx 28 mins]..
[audio:http://bradblog.com/audio/KPFK_BradFriedman_RogerJohnston_ArgonneRemoteDieboldHack_092811.mp3]
P.S. There’s been a fair bit of follow-up coverage of my Salon article yesterday, here’s a few of the better reports I’ve seen, at The Register, ComputerWorld, CNET and MSNBC.com.
Note: Of the media outlets listed above, only CNET got the ownership of Diebold voting equipment correct in their reports. Both The Register and ComputerWorld incorrectly report that ES&S now has ownership of what had been Diebold Election Systems, Inc. In fact, while ES&S, the world’s largest e-voting company, did purchase the company (that some refer to as DESI, but I do not, for personal reasons) in late 2009, the DoJ ordered them to sell it off due to anti-trust concerns. It was then purchased by Canadian firm Dominion Voting, though ES&S was allowed to retain a fair number of the contracts from Diebold jurisdictions that preferred to stay with ES&S. So while Dominion officially owns what was Diebold’s election division, ES&S still maintains and services many Diebold machines as well. Dominion also took over Sequoia — and lied about it — shortly thereafter.
Thank you Brad for your tireless efforts to get the truth out about election fraud.Your post on DU got over 100 comments..and a lot of them showed so much knowledge about the facts.
Here in ME , our Tea Party governor is trying to stop registration on election day.I have given your link many times on online comments in the Bangor Daily News.The only newspaper of any size for middle and upper state ME.
Thank you for always remaining factual and courteous. Know it hasn’t been easy.
Good post.
Let’s face it, if they can fly drones over difficult terrain from across the planet, using a “joy” stick, to find a target, one of these voting machines is elementary to hack.
What we can’t seem to do, which is more difficult, is to get through the status quo psychosis layer in the collective mind of the populace to convince them that this is actually happening.
Brad,
I haven’t listened to the interview yet. But don’t you think you ought to clarify that what Johnston thinks could be hacked are the reported results from the optical scan machine? The paper ballots would still be there for recounting or audit, given proper chain of custody (which is always a big if).
I don’t know of anyone who has ever claimed that computer-based optical scan tabulations cannot be hacked. The big advantage of paper ballot optical scan over ATM-style voting is INDEPENDENT CONFIRMABILITY, not inherent security.
We disagree about the minimum standard for independent confirmability. You insist only hand counting can work. I believe making digitally signed ballot images public would be a huge step forward, and enable a wide variety of approaches to confirmability.
I agree with you that hand counting has advantages. I just don’t think the nation’s election officials are ever going to go there, and I think it introduces its own issues.
I believe making digitally-signed, verified images public is a step many elections officials will eventually take, either voluntarily or, if need be, as a result of court order(s).
Dr. Roger Johnston’s assertion of opti-scans might change if had the opportunity to play with one.
Is Diebold (whatever their current names is…) willing to prove that their opti-scans are secure by lending out a machine to Argonne?
Since there’s NO chance key people in the US Dept of Justice and at the nation’s major news media are ignorant of this issue, I’d say it we should safely assume they’re are all in on the joke, and have been all along.
Still, I just don’t see how they can control ALL of national news media. Have they really convinced everyone a news blackout is a national security issue?
Maybe it’s not so crazy… They’ve shut down most all coverage of the Wall Street Occupation.
Just seems to me its their personal security that’s at risk here, not the nation’s.
FUN FACT: Our election commission chairman used to dress in a gorilla suit and roller skate the corridors of the county building in DuPage before he was appointed to this position.
Keep that image in your brain whenever you bring up DuPage elections, okay?
Mitch Trachtenberg @ 3:
I’m pretty sure that’s exactly what I did. Here are the two different references to Johnston’s comments about opt-scan hacks from the above article…
…
the hack low-rent attack Argonne has demonstrated …could also be implemented … on the paper ballot op-scanners used to count the majority of votes that will be cast in the U.S. in next year’s Presidential election.
Are you reading the above differently than me??
Well, voting machine companies and election officials aside, of course…
True. In the same way that suicide by overdose offers a big advantage over a gun to the head, because maybe the poison can be removed if someone bothers to stomach pump later.
Overstatement? Yes. Way. But hopefully my point is understood. If nobody bothers to “INDEPENDENT(LY) CONFIRM”, and in a way that guarantees the chain of custody is secure (a “big if”, as even you noted), that “big advantage” is almost non-existent.
To be clear, I do not “insist” that. But in nearly 10 years of doing this, and seeking “solutions” to whatever the “problem” is that we have here, I have found nothing more transparent (and thus, verifiable) than precinct-based hand counts. I don’t know if it “can work” everywhere or not, as I’ve called for pilot programs everywhere to find out. However, it does seem to work very well wherever it is actually used.
We should offer some disclosure here: You have created a very good system for doing exactly that, and hope to see it adopted around the country. (Not using that point against you here, I just think disclosure will help inform others following along).
Your system seems to be a good one, and certainly offers more transparency and independent confirmability than any non-hand-count system out there, that I know of. So I’m in favor of it in that regard, while having a number of reservations about it that I’ve expressed to you in the past (and needn’t go into here, for the moment, unless someone requests it.)
I’m not big on the “never gonna happen” logic for anything. If it’s the right thing to do for our country, then the right thing to do is continue advocating for it until it happens. Successfully or otherwise.
But I’m curious as to what you feel are “its own issues”?
They may. And I support your right to advocate for exactly that. I also support the “right” to advocate for plans which endanger our democracy, like secret vote counting and Internet vote counting, and Instant Runoff Voting and Vote-by-Mail, etc. But unlike many who support those bad ideas, I appreciate that you are truly advocating for what you feel is a good idea for citizen-overseeable democracy. So thank you for that!
FayWray @ 4 said:
Bwahahahah… Oh, you were serious?
Brad,
The overriding issue for hand counting is being able to staff precincts with sufficient numbers to hand count every race accurately and reliably at the close of polls.
If hand counting is delayed until after ballot movement, it is subject to chain of custody issues, perhaps even more so than machine voting where results can be noted at the precinct.
So what you’d need to have for hand counting is a team of at least four (reader, witness, two tabulators) with representatives from multiple parties, at every polling place, willing to stay until they’ve reliably counted every race on every ballot. On a large ballot, that might mean until daylight. (People find that hard to believe, but they should watch the hand count process done on a sample precinct as part of standard California election procedures, or at least listen to someone who has participated in such a count.)
And then you’d need rules in place to ensure that the four people could not collaborate to defraud an election.
If a law requiring hand counting came up on my ballot, I’d vote for it, and I’d want to see the people trained well and paid well. Do I see it as happening? Sorry, I don’t. I do see, just barely, the possibility that (someday) scanners belonging to an independent agency might be placed at every precinct, with the resulting scans forwarded to that agency by a path independent of the elections office.
You can scan all ballots in less time than it would take to hand count a single race, and you’re left with the ability to have multiple unaffiliated groups count off the images. The elections office is then left with the task of explaining any discrepancies.
The system in Humboldt was developed precisely because some of our local election integrity folks recognized that there were legitimate logistical concerns with hand-counting.
I am the Democratic Election Commissioner in Columbia County, New York. Our county fought the good fight against using optical-scan voting machines, but a federal judge essentially forced us to. So we implemented incredibly tight chain-of-custody procedures to secure and then transport, via bipartisan teams on election night, all ballots and other election materials to our county jail. There the members of another bipartisan team, each with a key to padlocks securing a single jail cell, receive ballots etc. and lock them up.
Ballots remain in that triple-locked cell (the sheriff, of course, has a key) until we have teams in place at the board ready to receive the ballots, ascertain the chain of custody, and then count the votes.
We’ve done this through three elections so far. It’s a considerable amount of work, but my counterpart and I have no doubt as to the security of our ballots, nor do we doubt the veracity or accuracy of our ballot counters’ results.
We’d like to try election-night precinct counts, piloting first in one or two precincts before ultimately mounting a full-county operation, but I’m not sure when we’ll be able to do that. In the meantime, given the way we conduct our elections, we have a great deal of civic participation in the process. I don’t think there’s anyone in Columbia County who doubts the accuracy of the vote count that my counterpart and I ultimately certify.
Virginia Martin
Democratic Election Commissioner, Columbia County, NY
Following your exchange with Mitch Trachtenberg, I’d like to say that good post-election audits are practical as shown by the few states who use them. Chain of custody must be provided, but even rudimentary chain of custody requires a herculean effort to defeat in a way that would matter to all but the very closest elections.
Paul Stokes @ 11 said:
“Practical” doesn’t equal “what democracy requires”. Neither does it mean “effective”.
Fortunately, we don’t have very many of those, right Mr. Gore, Mr. Kerry, Ms. Kloppenburg, Mr. Miller, et al? 😉
Virginia Martin @ 10:
I cannot EXPRESS how great your comment made me feel when I read it late last night. Thank you! (And your GOP counterpart!) Please keep going! And please let me know how I can help! (As well as keep us up to date on your plans for a precincty-based hand-count pilot program!)
Also, just so I understand, when you say:
…Are the votes then counted by central op-scan? Or by central hand-count?
Mitch Trachtenberg @ 9:
That is certainly an issue that needs to be answered to. (And just one of the reasons my call has been for precinct-based hand-count pilot programs, as opposed to hand-counting outright. I don’t think it wise to make the same “mistake” the e-vote companies did when they said “Trust us! This will work perfectly!”)
That said, the places where it is done seem to have no problem doing it, and an abundance of folks turning out to help — as in 40% of the towns in NH.
Also, please note Virginia Martin’s inspiring comment above, where she says: “given the way we conduct our elections, we have a great deal of civic participation in the process.”
I hear the same thing from all of the folks in jurisdictions where they do that sort of process. Again, look at the inspiring turnout and civic pride and participation you see in NH! A few examples…
Agreed. (Well, dunno if I agree with “more so”, but I take your point.)
Right. And with a fresh team coming in to do it at the close of polls (let’s say 8p or so), to replace those who have been there for about 14 hours, pulled from a much larger pool of folks than we have to pull from for regular poll workers (since it’s after the work day), and with Ds worry that Rs will cheat if they’re not there and vice versa, it seems there should be little prob getting those folks and many more.
Then that means you simply don’t have enough counters, or your precincts are too large! (Making them smaller, btw, also has the advantage of decreasing long lines for voting too!)
Right. Though doing it all out in the open, with ALL parties and ALL citizens and ALL video cameras rolling, it’d be tough to get away with much chicanery. At least not without getting caught (and hey, you could even run the ballots through the Trachtenberg Independent Op-Scan method later in the week if you really wanted to, just for safety!)
Wonder if Virgina Martin above sees it the same way.
Again, for full disclosure, that is the plan you’ve developed and for which I see a number of problems, even as I support your efforts there.
Which unaffiliated groups? Americans for Prosperity? ACORN? SEUI? The Heritage Foundation?
While I still appreciate your thoughts there, I don’t yet see evidence that the “legitimate logistical concerns” would be realized. So far, the contrary has been shown to be true in places where precinct-based hand-counting is done.
BTW, I’m guest hoping Mike Malloy Show again next week, and hope to be speaking with NH’s Nancy Tobi, author of Hands-On Elections: An Informational Handbook for Running Real Elections, Using Real Paper Ballots, Counted by Real People! That’ll likely be next Thursday night! Hope you’ll tune in!
Brad,
I believe your last question/point is critical.
You ask: “Which unaffiliated groups? Americans for Prosperity? ACORN? SEUI? The Heritage Foundation?”
The answer is, “yes, all of the above, as well as BradBlog, The New York Times, and Mitch’s Vote Counting Shoppe.”
Ballot publication offers the possibility of millions of eyes bringing as yet unknown techniques to the detection of anomalies in ballots. (If, and it’s always the big if,there is satisfactory proof that the images truly represent the ballots.)
That has the potential for becoming a substantial deterrent to fraud, particularly if a technique relying on ballot images was able to catch a genuine fraud.
It’s a crowd-sourcing technique. Yes, validating the images would be a major concern. But I am convinced that this is a far easier problem to deal with than faking the numbers. More data will work on the side of integrity.
per Brad Friedman @13:
[I said] Ballots remain in that triple-locked cell (the sheriff, of course, has a key) until we have teams in place at the board ready to receive the ballots, ascertain the chain of custody, and then count the votes.
[Brad said]…Are the votes then counted by central op-scan? Or by central hand-count?
Brad, the votes are counted by central hand-count. All of them. And the hand-count numbers are the ones we certify.
Two more things:
The sheriff does NOT have keys to the Democratic or the Republican padlocks.
The New Hampshire voting process is amazing and inspiring!