Hacking an Election, Really?
Watching the news last week and over the weekend it is almost humorous to see the spin going on about the voting system issues (Sequioa AVC Advantage). It is almost discouraging people from trusting the voting system and technology all up. it concerning for me as an architect to see this happen. I am wondering if there was any architecture efforts from the either the client or the vendor side on this one. The issues seem so elementary.
Just to get it out of the way real quick, this is not a post about any sort of political statement, endorsement, attack on the media or "the man" for that matter. If you want to hear my personal views on politics friend up with me on twitter.
So if you haven't heard, the folks at Princeton University Center for Information Technology Policy did a security assessment on the electronic voter systems and found some really big security flaws in the voter systems that are deployed in several states including key swing states like Ohio. This was brought on by a New Jersey lawsuit brought against the state by public-interest groups. The full article can be found here: http://arstechnica.com/news.ars/post/20081024-study-sequoia-e-voting-machines-disturbingly-easy-to-hack.html and http://freedom-to-tinker.com/blog/appel/report-sequioa-avc-advantage
Andrew Appel from Princeton University who led the study says:
"We have found that the Advantage AVC firmware has errors. We have also found that it is easy to replace firmware in the AVC Advantage with fraudulent firmware that can undetectably steal votes and thus change the outcomes of elections," the report says. "Furthermore, some kinds of fraudulent firmware can automatically virally propagate themselves from one AVC Advantage voting machine to another, without the attacker being physically present. Once fraudulent firmware is installed in the AVC Advantage, it can steal votes in election after election without any additional effort by the attacker."
After talking with family and friends that are not technology savvy it was interesting to hear there responses to this bit of news. There was a great deal of frustration with the vendor, government agencies that approved the machines and technology in general. Is this fair? Who's to blame? And who is ultimately on the hook for this? All of which are hard to answer with any real certainty.
As a technologist I have to take a step back and look at the broader picture here.
Risk vs. Probability
While the risk is high, the probability for this sort of incident is low. To pull this off you really need someone that understands technology and is comfortable with modifying a chip set. I would think someone would have to do the following:
- Read up on DYI Guide to hacking a voting machine
- Get the tools & software to flash a chip. This would require some low level understanding of computers which is going to be a relatively small portion of folks out there.
- Possibly get a chip that is already loaded, now they have to purchase somewhere. Online? The mysterious guy in an ally somewhere? You get the point.
- Then you have to go to register, put your name on a piece of paper, go into a booth and hack away
- If you can get to the hardware components and if your handy lock pick or screw driver doesn't make too much noise you may be able to pull it off
- Pull and pray... What if it doesn't work? What if the system goes into an endless beep cycle when reboots... Well, your toast.
There are way too many points in this process where you as the hacker are personally identifying your self. I really don't see why any smart person our there would do this.
Risk vs Reward?
So what's in it for the hacker? Is it the thrill of "putting it to the man"? Is it trying to get your favorite politician in office, like in your favorite episode of Heros? As I outlined above, there is a ton of risk here for arguably little reward. The consequence, breaking a few federal laws and spending a good portion of your life in "the clink".
I don't see the incentive.
Ability to do this?
As I mentioned above, to pull this off successfully there is a requirement to be very technology savvy. While there are many of us out there, I am confident we are using our skills for good rather than prison time.
So what does this mean?
In my personal opinion, at the end of the day whomever purchased these machines should be ultimately accountable. They did not due the required due diligence. I am somewhat disappointed in the lack of diligence here given the federal mandate to consolidate technology decisions, create structured processes and enterprise architecture. I wonder where process broke down on this one?
As for the blame on technology, I don't think it's a valid argument. We trust technology with our money, our health care and many other critical things in our life.
So what about the news media. Are they to blame for anything? Well this is more of a philosophical debate but personally they are in it to make a buck and get higher ratings so just like with anything they are going to hype it up. While I don't agree with the methods, I understand them and have my hype filter set on high.
As for the vendor, there should be some accountability there since they provided the software. It more of an ethical dilemma really. Ultimately the government purchased this horribly flawed machines the vendor should of been forced to make changes on the behalf of it's customer. Personally they probably should of stepped up a bit more since they knew well in advance what these machines would be used for.
To sum it all up, can I get some more architecture here? After looking at the entire solution there seem to be a number of issues besides security. One important one was the user experience (UX) aspects. The UX was pretty bad and not all that functional or easy to understand.
- Insecurities and Inaccuracies of the Sequoia AVC Advantage 9.00H DRE Voting Machine (click here)
- Frequently Asked Questions
|Share this post :|