An scary article on Ars Technica about how insecure the new voting machines are:

How to steal an election by hacking the vote

From the article:

In researching this article, I talked on- and off-the-record with a number of prominent experts in the electronic voting field. The following e-mail response from computer scientist Peter Neumann sums up the present state of chaos heading into the November midterm election, and it also communicates some of the frustration (and fear) that I heard echoed in the responses of the other researchers whom I questioned.

"The problem is much deeper than most people realize. The standards are extremely weak (1990 and 2002 both), and VOLUNTARY. The systems are built to minimum standards rather than attempting to be meaningfully secure. The evaluations are commissioned and paid for by the vendors, and are proprietary. The entire voting process consists of weak links—registration, voter disenfranchisement, voter authentication, vote casting, vote recording, vote processing, resolution of disputes (which is essentially nonexistent in the unauditable paperless DREs), lack of audit trails, and so on. You cannot begin to enumerate the badness of the present situation."

Paradoxically, the media blizzard of disparate facts, figures, vulnerabilities, acronyms, and bad news from a huge list of states, counties, and precincts, is in large measure responsible for the current lack of an all-out panic among the public and political classes as we head into the November mid-terms. This steadily roiling storm of e-voting negativity has resulted in a general uneasiness with DREs among the public and the media, but the threat feels diffuse and vague precisely because there are just so many things that could go wrong in so many places.

-Chris
Studentfilms.com