December 15, 2003
Guess what? There's an electronic voting machine controversy!

There may be a recount in the City Council District G election, where Pam Holm won by a 27-vote margin over Jeff Daily. The timing is perfect, coming on the heels of yet another report that electronic voting machines, including the eSlate machines used in Harris County, have security flaws.


The analysis, conducted for the Ohio secretary of state, of Hart InterCivic and three other vendors' systems found 57 potential security problems. Hart InterCivic, Harris County's eSlate vendor, had 10 potential risk areas, including four rated as high.

"We believe because of weaknesses we found in all of these systems, the vendors need to go back and take care of the weaknesses," said Glenn Newkirk, president of InfoSentry, one of two firms hired to conduct the review.

[...]

Researchers from InfoSentry, of Raleigh, N.C., and Detroit-based Compuware Corp. reviewed Hart InterCivic, Diebold Election Systems, Election Systems and Software and Sequoia Voting Systems. The six-week review was prompted after security concerns were raised in Ohio and elsewhere.

The review found, among other things, the potential for an unauthorized person to gain access to eSlate's supervisory controls and shut down the polls early. A password is required to shut down the system here. Hart InterCivic contends that if the system were shut down, voting data would not be lost.

The report also notes that eSlate lacks encryption to protect voting data, and Hart InterCivic is now considering the change.

Another risk identified in the report is that the connection between the system's units can be accessed by voters and disconnected. More security would alleviate the risk, the report states.

Results in which security breaches failed include: an unsuccessful attempt to access the system from an external source, failure to load a program through external sources and failure to upload results twice.

"Compuware has identified several significant security issues," the report states of the eSlate system, "which left unmitigated would provide an opportunity for an attacker to disrupt the election process or throw the election results into question."


The fact that the testers were unable to fully hijack the eSlate system is a point in its favor, but when I read things like "lack of encryption", I have to wonder what Hart InterCivic was thinking when they designed the system. All we've really got here is assurances from the vendor and the County Clerk that we shouldn't worry our pretty little heads over all of this, and they both have strong incentives to take a sunny view of things. Whether it happens in this recount or not, I believe it's just a matter of time before we get bitten on the ass. When we do, I hope everyone remembers what the responsible parties have been saying all along.

UPDATE: Rob actually read the report, and has some good thoughts on the subject. Check it out.

Posted by Charles Kuffner on December 15, 2003 to Election 2003 | TrackBack
Comments

This is a somewhat misleading article (a shocker in the Comical, I realize). It gives the impression to a casual reader that problems with e-slate have prompted this potential recount, when the primary contention seems to revolve around the signatures required to obtain a ballot on election day. And honestly, the main thing driving the potential recount seems to be that one candidate is disappointed in losing a close election.

Posted by: kevin whited on December 15, 2003 8:20 AM

For those interested in the problems inherent in electronic voting technology, Bev Harris is going to be holding a press conference to release new information at 2pm tomorrow. Ms. Harris is the author of _Black_Box_Voting_ and the woman most responsible for the investigation of the flaws in Diebold's systems, etc. She hasn't yet revealed what the press conference is about, but apparently she's uncovered new information regarding the unreliability of these systems, and judging by her past conduct, she doesn't get this excited for no reason. If she has information, I'm sure it is extremely thoroughly researched and well-documented.

No information on who is covering the announcement, or where/whether the information will appear on mainstream news, but you can keep on things at blackboxvoting.org

Posted by: Jeremy Robinson on December 15, 2003 12:52 PM

I took a look at the report and commented on it here.

Posted by: Rob Booth (Slightly Rough) on December 15, 2003 2:53 PM

I have to agree that electronic voting machines are unreliable. I have first hand knowledge of their unreliability (I wish I could say more, but I can't).

This shouldn't be a partisan issue. I'm unsure what is better (punchcards? heh), but I'm firmly not convinced that e-voting is the way to go (and even more convinced that internet voting is NOT the way to go).

Posted by: Another Rice Grad on December 17, 2003 8:08 PM