Voting machines only failed in our strongholds, says Azimio witness

Kenya Integrated Election Management System (KIEMS) Kit, on August 9, 2022, during the General Election. [Edward Kiplimo, Standard]

The voter apathy witnessed during the General Election on August 9 was by design, Azimio la Umoja-One Kenya Coalition presidential candidate Raila Odinga and his running mate Martha Karua have now claimed.

In the presidential election petition filed before the Supreme Court, Raila and Karua allege that the Kenya Integrated Election Management System (Kiems) only failed in their strongholds to ensure their supporters left the polling stations without voting.

The petitioners' witness, Benson Wesonga, claims that although the Independent Electoral and Boundaries Commission (IEBC) had guaranteed Kenyans that it would ensure all voters would be identified through the electronic system without any hitches, voters in Kakamega and Makueni had to ultimately use the manual register.

Mr Wesonga, who is an IT expert, said the decision to revert to printed copies of the voter register was "too late in the day" and residents left without voting. He cited at least 210 affected polling stations in the two counties.

"The commission willfully and negligently caused the failure of Kiems kits in polling stations in Kakamega and Makueni counties-perceived as the petitioners' strongholds-in order to suppress voter turnout in the said counties.

"The commission is on record advocating for the exclusive use of Kiems kits ostensibly as part of a wider scheme to suppress voter turnout that to stymie abuse of the manual register. There is reasonable cause to believe that the failures were deliberately orchestrated to suppress voter turnout," says Mr Wesonga in court papers filed on Monday.

He also questioned why IEBC had forms 34A in PDF form on its portal while its advertisement looking for bidders had allegedly indicated that the software installed in the kits ought to convert the still photos taken into JPEG.

According to Wesonga, having PDFs instead of JPEGs in the public portal is an indicator that there was a human hand between the Kiems kits and the servers.

He alleges that the manipulation is the reason why other interested parties were unable to come to a conclusive tally on who won.

"I am aware of my own knowledge that the above process can be manipulated through what is commonly known as a 'man in the middle' attack or 'machine in the middle' attack in which a person authorised or unauthorised can intercept and manipulate data.

"Plainly put, the results were recalibrated to achieve a particular pre-determined outcome. The foregoing is compounded by the fact that the media houses, the general public, and the four commissioners of the IEBC were unable to determine the results of the 2022 presidential elections at the date of declaration of the impugned results, and to date, there appears to be no consensus as to what the results are really are."

Wesonga gives the example of Kagera Primary School polling station which had two form 34As in the portal; one showed Deputy President William Ruto had garnered 115 votes while the second one showed he had 195 votes.

Other stations with two form 34As, he says, are Borut Primary, Mugumo Primary, Thumaita Primary, and St Martin's Kibagare. Wesonga's court papers read that at Borut, one form in the portal indicated that Ruto garnered 109 votes while the second one says he had 169 votes.

Thumaita form, according to him, has several entries in the same form uploaded to the IEBC portal, including the number of agents who signed the forms. He says the polling station count section is blank and the handwritings are different. The agents' signatures are also not the same.

"The above samples prove two things. First, that contrary to the commission's assurances, the forms were susceptible to and were in fact converted to PDF specifically to facilitate manipulation; and secondly that the results transmission system was corrupted and was so compromised that all the forms submitted therein cannot verify the physical results that were submitted to the national tallying center," he says.

Wesonga claims that it is possible to manipulate figures without necessarily needing to recall agents or returning officers to sign the forms afresh. He adds that a cursory view of forms 34B supplied by the commission at Bomas of Kenya for Nakuru, Kiambu, Kisumu, Meru, Kakamega, and Mombasa showed that the commission's votes had been inflated by over 180,000 votes.

"It is absurd and statistically abnormal that errors in the transposition of results would be in the thousands and only in favour of one candidate," he says.