ATLANTA GA -- An in-depth Root Cause Analysis released today documents how voting equipment security flaws led to the human error that created delays and controversy in producing the 6 th District special election results on April 18th . The analysis revealed two "critical security flaws" and two "serious security flaws" that span Georgia county servers, voting machines and electronic poll books.
Most of the security flaws involved the software's failure at each level to detect an incorrect voting card from a Roswell runoff that was entered into the 6th District county vote totals. Fulton County had to use redundant databases and procedures for simultaneous parallel elections on the same day to accommodate three elections scheduled on April 18th . The analysis cited inadequate error detection in the software and improper election scheduling as another root cause of the glitch.
When the problem was corrected, leader Jon Ossoff's vote totals dipped below 50%. That caused a runoff with second place candidate Karen Handel who was nearly 30 points behind at 19% of the vote. The election reporting snafu and runoff led to national skepticism about the outcome. Georgia has been under national scrutiny since 2002 when it implemented what national experts call "unverifiable elections". The phrase denotes the equipment's lack of an independent audit trail.
The analysis produced by VoterGA concluded that, "The GEMS servers are vulnerable to fraud and critical errors that can dramatically alter the results of an election". It stated, "The GEMS databases at any county will accept invalid election data from a remote or local source and inject the invalid data into live election results for totaling and publishing."
Since 2002, hundreds of critical and serious security flaws have been documented by a variety of studies examining the same type of voting equipment that has been used in Georgia for 15 years. These studies have been conducted by universities such as Stanford, MIT and Johns Hopkins, technology firms such as Compuware, SAIC and Raba Technologies and states such as Maryland, Ohio, Nevada and California. The machines have a useful life of about 10 years.
The VoterGa preliminary analysis did not find evidence of fraud in the elections but it explains how fraud could be undetectable. It was produced with input from national voting experts and technology professionals in Georgia and other states.