A bipartisan group of software engineers has come out with initial findings from the 2020 elections in Michigan showing anomalies suggesting a computer algorithm was functioning inappropriately in voting machines.
The group was led by Shiva Ayyadurai, an inventor and software engineer who recently lost the U.S. Senate race in Massachusetts, Bennie Smith a Democratic Party elections official in Shelby County, Tenn. and Phil Evans, who is a software engineer and entrepreneur.
The group's findings were clear, the election integrity in many jurisdictions across the country is vulnerable.
“This is not a Republican, Democrat, independent issue – this is an American issue,” Ayyadurai said. “It's an engineering issue.”
In Michigan, the group said 2020 election results from four counties they studied appear to show evidence of an algorithm improperly influencing the outcome of the voting machines tally. One of those counties was Oakland County.
In particular, the group pointed to a feature on voting machines that allows for votes to be weighed. In other words, the vote loses its value as a single vote once entered into a voting machine.
“The documented feature of a weighted election is such, such a bad idea,” Smith said. “The only thing that can verify to us in a black box is the input and if we don't know, we don't know. We can't know if a weighted election was conducted. It's a feature, it's not a bug. So the only way we could get back to that is for the citizens to be able to remove the computers from the situation.”
The group said they believe Trump votes were transferred to Biden and that their analysis backs up such a claim, but that there simply is no way to properly verify the election.
In Oakland County, Ayyadurai said Evans looked at data for the county and found that it looks like 20,000 votes were taken from Trump and given to Biden.
Ayyadurai said once Trump hit 20% of votes, it looked like his votes became linearly reduced.
“You find out about 20,000 votes were taken from Trump and 20,000 votes were given to Biden for 40,000 Biden lead in the Oakland early voting, just one county,” Ayyadurai said. “So that's what you see here.”
Ayyadurai said below the 20% line, Trump was getting more votes than other Republican candidates, but right at 20%, Trump’s vote counts started getting linearly reduced.
Ayyadurai said in data and pattern analysis or if you’re in AI, you look for features similar to this.
“So this beautiful curve, it's flat and boom, it drops in a nice linear fashion,” Ayyadurai said. “As though an algorithm was there as you went across, votes were getting cut by Trump, and that's what we see here.”
Ayyadurai said the pattern is the same with both Election Day votes and early voting.
“Trump's votes are reduced by 10,000 and Biden's votes are increased by 10,000 here,” Ayyadurai said.
Ayyadurai said their position is that the linear decline is so precise over so many precincts that it could only be done by a computer algorithm.
Alerts Sign-up