Dominion Image Cast X (ICX) voting machines have been tested in a Georgia lawsuit by a computer science and engineering professor JR Halderman who has made a career studying electronic voting security. The Professor issued Security Analysis of Georgia's ImageCast X Ballot Marking Devices, also known as The Halderman Report in July 2021 identifying nine critical national security vulnerabilities, including malware which can be installed by remote access while subverting all security procedures and Logic & Accuracy testing. These voting machines are utilized in 12 Ohio counties: Adams, Butler Fairfield, Greene, Hancock, Hardin, Perry, Richland, Scioto, Stark, Wayne, Wood.
The Dominion ICX voting machines print a cash register type receipt known as the Voter Verifiable Paper Audit Trail. Votes cast are printed in English text on the receipt, along with a QR code. The problem lies wherein the voting machine reads the QR code to count the votes, not the English text on the receipt. The voter really has no idea if their vote is counted as desired, which is against federal law.
The following is a hypothetical example of an election in Ohio that changes a candidate from being ELECTED to a candidate being SELECTED all the while making it virtually impossible for a post-election audit to detect the cheat.
130,000 votes cast in an election in which candidate A is Elected
65,260 (50.2%) votes were cast for candidate A
64,480 (49.6%) votes were cast for candidate B
260 ( 0.2%) votes were cast for candidate C
In the example, malware is installed by remote access and changes 975 (0.75%) votes for candidate A to candidate B. Even though the elector sees candidate A printed on the text of the Voter Verifiable Paper Audit Trail, the QR code that also prints on the Voter Verifiable Paper Audit Trail is created to indicate a vote for candidate B.
Can you read QR code?
129,025 votes remain the same, but now candidate B is Selected, not Elected. Results after malware infiltration:
64,285 (49.45%) votes for candidate A
65,455 (50.35%) votes for candidate B
260 ( 0.2%) votes for candidate C
No worries, the post-election audit will catch this, right?
5% of the 130,000 total votes are audited, which is 6,500 votes. 123,500 votes are not audited. It only took 975 votes to flip the election.
The odds that 1 of the 975 votes changed are included in the 6,500 total votes audited across the county in which 130,000 total votes were cast is less than 1%.
What about a full hand recount of the votes?
An accurate full hand recount would catch the 975 votes for candidate A that the voting machine changed on the QR code as candidate B. At least 13 of the 975 votes changed must be audited to attain less than a 99.8% SOS defined post-election 'accuracy rate' to qualify for a full recount, which is virtually impossible. Therefore, a full hand recount likely will never happen, the 975 changed votes will never be detected, and candidate B will be SELECTED as the winner of the race.
l