Courts have settled the law on presidential election results: once openly collated, announced and certified by a constituency returning officer, those results are final and can only be challenged through a petition.
The remaining national task is to collate the results from the 290 constituencies and declare the winner. That is what the judgment in the Maina Kiai case says. Technology was introduced to make elections more credible. It was supposed to make voting, identification, transmission, tallying and verification more accurate and transparent. Instead, after nearly two decades of experiments and billions of shillings spent, elections have been turned into an increasingly inaccessible black box.
Over time, the election server has become the most mysterious and politically consequential part of every presidential poll. Fungua server—open the server—has become a familiar post-election refrain in Kenya’s political discourse. It is the technological equivalent of “show us the ballot papers”, except that a ballot paper can be held up, examined and counted. The servers have been examined, but never in a way that has settled the political argument over what happened inside them.
This week, a familiar dispute broke out over who should supply the election machines. A Sh6.5 billion contract for an Integrated Elections Management System has touched off a procurement challenge over claims that its specifications favour South Korean firm Miru Systems, while Smartmatic, which supplied Kenya’s electoral technology in 2022, remains part of the wider contest over who should provide the machinery of the next election.
Kenya’s history with election technology is not encouraging. In 2007, Kenya acquired election technology but failed to deploy it. In 2013, biometric voter registration, electronic voter identification and electronic transmission of results were deployed. When the systems failed, the election reverted to the manual processes technology had been expected to replace. In 2017, the technological ambition had grown further, yet technology became entangled in the crisis it was supposed to prevent. The Supreme Court annulled the presidential election because the IEBC had failed to conduct it in accordance with the Constitution and the law. Among the problems was the failure to properly transmit and verify results electronically. Thousands of polling stations had not transmitted results when the presidential vote was declared.
Technology had been sold as the answer to the credibility problems exposed in 2007. Instead, failures surrounding its use became part of the reason the 2017 presidential election had to be repeated. Technology reportedly worked better in the last election. The electronic transmission system functioned and the IEBC published results from polling stations. But the technology did not settle the question of legitimacy.
The Supreme Court ordered scrutiny of the electoral technology, but the manner in which the IEBC facilitated that scrutiny remains a sore point for the petitioners’ legal team. The IEBC reportedly provided only part of the documentation it was ordered to. More importantly, the petitioners’ lawyers said they were not granted access to the authentic servers used in the presidential election, making it impossible to validate the documents provided. When asked for the forensic image of the Results Transmission System, the IEBC invoked the intellectual property rights of its technology supplier: Smartmatic. These concerns expose the central contradiction: A system can be technically secure and still be democratically opaque. Credibility cannot be engineered into a machine.
Buying biometric kits, servers, transmission systems, cybersecurity and a new integrated election management system does not translate into credibility. Credibility comes when voters can trust the institution conducting the election—and can see enough of what it does to know why they should trust it.
The more complicated the technology becomes, the more citizens must take on faith what they cannot see. Technology was supposed to provide a clearer sight of how the vote was counted. Instead, it has become a blindfold.
The electoral system is required to be simple, accurate, verifiable, secure, accountable and transparent. Yet the technology scrutiny surrounding the 2017 and 2022 elections showed how difficult it can be to establish whether those standards have actually been met.
The question is not whether Kenya’s machines are more sophisticated than they were in 2007. It is whether the citizen is in a better position to verify the election. On that measure alone, the progress is far less obvious.
Technology is not cheap. The IEBC has asked Parliament to provide Sh1 billion for results transmission for the 2027 election, alongside Sh3.8 billion for systems upgrades, and Sh6.2 billion to replace Kenya Integrated Elections Management Systems kits. Every election now comes with an added technology bill: new systems, upgrades, replacements and another round of procurement. What democratic return are they getting for the money? Is spending Sh1 billion on results transmission technology necessary when the law already says the presidential election results are final at the constituency? The promise was to give Kenyans greater confidence in the vote: accuracy, transparency, verifiability and accountability. Not better machines. If the constituency result is final, perhaps technology is being asked to do more work than the Constitution requires.
The safeguard should be much simpler: by the time the result reaches Nairobi, it should already be visible to Kenyans at polling-station and constituency level. The demand to “fungua server” is not merely the cry of a disappointed loser; it expresses a democratic demand to see the machinery by which the result was arrived at. The IEBC should be careful about making the black box more expensive.
The writer is a board member of the Kenya Human Rights Commission and writes in his individual capacity. @kwamchetsi; [email protected]
