The battle for Kenya’s next president is not being fought only at rallies, in televised debates, or in newspaper headlines. It has slipped into the quiet corners of daily life. It lives on the phone screens of day laborers, market traders, university students, and shepherds in rural villages. A woman in western Kenya scrolls through a forwarded video and sees a political leader supposedly giving money to a crowd. A farmer in the Rift Valley listens to a WhatsApp audio claiming that election officials are planning to rig ballots. A young voter in Mombasa sees a photo of a candidate shaking hands with a shadowy figure, and the caption says everything that need not be said. These posts travel at breakneck speed, and the language they speak is not evidence but certainty: “Nimeiona kwa social media.” In that phrase lies the new crisis of democracy. Because once a piece of information appears on a phone screen, many Kenyans now assume it must be true—no matter how bizarre, no matter how old, no matter if the image has been doctored, or the video was shot years ago in another country. The story spreads because it feels real, because it provokes fear, anger, or excitement, and because sharing has become a reflex rather than a judgment. Kenya’s next election, then, will be fought as much on the terrain of public trust as on the terrain of votes. It is easy to imagine that democracy is threatened by the winner who loses, or the losing who refuses to accept. But the larger danger may be whether anyone knows what is actually true. If a nation goes to the ballot box confused, divided, and convinced by a million crafted lies, the freedom to vote becomes a very distant mirror. The facts are not the only casualty. The shared belief in a common reality—the very foundation of democracy—is what begins to break down.
Para2:
Kenya has seen this before. In 2017, researchers uncovered coordinated online campaigns designed to manipulate ethnic emotions, to paint one candidate as a foreign agent, to present another as the only hope of protection. Fabricated survey result, doctored images, and misleading videos flowed across Facebook, Twitter, and WhatsApp. People were attacked through manipulated histories of their communities. By 2022, the situation had deepened. Fact-checking groups, including Africa Check, Code for Africa, and several Mozilla-backed researchers, were almost overwhelmed by the flood of false claims surrounding election results. The names of the Independent Electoral and Boundaries Commission, its officials, the candidates, and even the rule books were all contaminated in the public imagination. Fake tally sheets circulated as if official. Old videos of violence from other years were reposted and labeled as recent attacks. Photographs of politicians were edited to suggest endorsements that never happened. Quotations were invented and attributed to leaders, giving them “quotes” intended to enrage or shame. The point was never simply to get people to believe the lie itself. The point was to irreparably pollute the information environment. When citizens hear contradictory accounts, when nothing seems trustworthy, when a legitimate result somehow not be believed, then the institution that Authority—the very machinery of the democratic system—begins to crack. The danger of fake news is therefore not only that it deceives a few people. It refuses to allow others to share the same sense of common realities. Researchers kept fact-checking, but the false stories went around quickly, with people passed them to their families and friends. The correction comes later to a smaller crowd, and a false first impression often holds. In that environment, confidence collapsed. The next election may expose that the same dynamic will be repeated, except this time with the tools of the tech giants and artificial intelligences that make the false videos, audios, and images even harder to distinguish from the genuine.
Now comes the most terrifying part: the arrival of generative artificial intelligence. It has drastically lowered the cost of deception. Previously, manipulating images and creating believable voices required skill, money, and time. Not now. Any person with a smartphone, a data connection, and freely available AI tools can produce a “deepfake” video of a presidential candidate at an event they never attended, or an audio recording in which a leader seems to be denouncing a whole community in crude terms. It is even easier to make a fabricated video of an electoral officer announcing imaginary results, months before voting is opened. The voice sounds real. The lips move naturally. The lighting is correct. There is no obvious tell for the ordinary viewer. Imagine reaching out to your phone at 6 a.m. and seeing this video: your candidate, whose voice you know from rallies, saying what you would never suspect, for example, that he is set to ally with a rival, or that he insults a ethnic community he has always respected. Even if a fact-checker later debunks it, the disgrace opens before the necessary correction. Corrections rarely travel as far or as fast as falsehoods; people remember the original headline or the original clip, not the deep paragraph buried in a correction. Political operatives know this. They have learned that memory is not just pictures—it is emotion. A deepfake creates a feeling, a reaction, a conviction that persists because of the internal circuit of disgust and fear. They are also learning that, instead of using gas propulsion to target, AI can flood the room with thousands of personalized messages. Voters can receive, all in one line, an attack tailored precisely to their own fears and loyalties. Micro-targeted propaganda becomes a new campaign strategy. One community may see a candidate’s commitment to land rights; another may be shown a video of the same candidate coddling land grabbers. Each group receives a different truth, apparently so personalized that it satisfies their existing distrust. In the process, reality is no longer a set of facts shared by all. It is a series of disjointed stories, each crafted for the largest debtor. Fact-checking is becoming slower, but propaganda is accelerating. To defend democracy we must now confront this commercial empowerment.
Why does misinformation win? The answer lies in the shape of digital spaces. Algorithms built to keep us online reward outrage, fear, anger, and shock. The conflict is what keeps a user watching, scrolling, sharing. The truth is often boring; misinformation is emotional, entertaining, and quick to share. The platforms have a financial benefit, because a user who stays and contributes to advertising revenue is a more valuable user. Long-term engagement, therefore, is a quality that platforms are designed to maximize. If the algorithm rewards outrage, then those who distribute lies are ultimately working with the platform, rather than against it. This is why so many false stories become visible without any further explanation: they are already spread by an algorithmic logic that has no notion of truth, only of engagement. At the same time, who is to responsible when misleading videos are generated by AI and go viral? The origin may be fake, built outside legal duty, hidden behind accounts that do not exist. A law can punish an individual who publishes hate speech, but how can it punish the anonymous mother who simply presses the “forward” in a moment of anger? How can a regulator distinguish a malicious deepfake from an ordinary political parody or gift-content? And how can the platforms be pushed, as former US President Joe Biden said in one of his farewell addresses, to protect their users from misinformation? Biden warned America is being buried under an avalanche of misinformation and disinformation. He said the free press is crumbling, editors are disappearing, social media is giving up on fact-checking, and the truth is buried by lies told for profit and power. His warning says more than what happens in HIS country. It is a pronouncement that has reached villages in Kenya. The abuse of power he speaks about is not confined to government: any individual with a smartphone and an impulse to share can become an abuser of the truth. That is why the answer cannot be a simple one-for-one, but a change in the public culture and in the responsibility of all who use, own, and regulate these platforms.
Protecting the upcoming election will require many people to act—not just government or social media companies. Governments need to make laws stronger without violating the rights to freedom of expression. Technology companies must moderate content during election periods more responsibly, instead of being active. Political leaders should sign ethical campaign codes and publicly reject the use of fabricated microphone, video and manipulative donation. The media, for its part, must invest more in a “verification-first” form of journalism. When an audio is viral, the newsroom should check before publishing; when a quote is quoted, it should see play and transcript. But all these efforts can be killed if ordinary citizens are not prepared to change their own emotion. A Nigerian voter doesn’t need to be an IT expert to be a simple citizen. Before he presses the forward button, stop. Ask: whom did I receive this from? Is it also reported by trustworthy news organizations? Or is it just cleverly made to make me angry or make me afraid? Maybe in silence, a good opportunity appears. Every text may not deserve a share. If a message makes you feel angry or safe quickly, that is a red flag. You need to pause, carefully check, in fact. In the internet world, the impulse to share/fast becomes stronger than the desire to be true. But the same impulse can be trained out of. When citizens begin to weigh evidence, they change the dynamics: false stories lose momentum because there are fewer willing carriers. This is not censorship; it is discernment. An informed public is the first line of defence. Not a fact-check, but the habit of questioning. The power to say, “Wait, let me look into that,” may be the most democratic act an old person can do today.
Kenya has spent three decades building its democracy. It has held multiple elections, faced shocks, improved institutions, and defended its civil society, independence, and independent media. It would be a tragedy of a high order if these achievements were undone not by an overt military intervention, but by something as cold as the algorithms and a viral fake video. The misinformation in Kenya is not just a delayed news story; it is a clamour in daily life, exploiting every existing weakness in trust, particularly along ethnic and social lines. If we allow the truth to be buried by an avalanche of misinformation, disinformation and fabricated content, we could risk losing something far more important than one election. We could risk losing public trust itself—the foundation on which democracy stands. But there is still a way to avoid that outcome. We can build an informed society that values evidence over anger and truth over comfort. We can teach ourselves to ask for the source and not for the emotion; we can model a civic life in which a correction is seen as a virtue, not a mistake. Biden’s warning is not a foreign worry; it is a warning for Kenya, for Africa, and for every democracy riding on digital age. The abuse of power has always existed. Now it can be hidden in a device in the hand of any one person. The “powers” are not merely in who controls state power, but in those who have a smartphone and a feeling of grievance. Therefore, the best defense of democracy is not censorship. It is not a crackdown that practically steals freedom. It is an informed and careful public, creating communities that are attached to true facts, but shows no use to serve falsehoods. Be the one who stops before sharing. Make the tiny decision not to amplify a lie. In that conscious pause lies a quiet, pleasing resistance. It keeps the space of public trust open, and it ensures that when Kenya goes to vote, the votes will be cast from a people who still know the real from the false, and who get to hear the truth.