Imagine it is 2027, and you are scrolling through your phone on the night of the general elections. In one WhatsApp group, a video is moving around: an INEC official, looking perfectly official, announces that a candidate has won a crucial state. In another group, there is a voice note in which a well-known politician seems to admit that a rival has already bought off electoral officers. Your heart races, and you share it. You are not careless; it just looks and sounds real. But none of it is true. According to a report by LEADERSHIP Weekend, this is exactly the scenario that election experts fear ahead of Nigeria’s 2027 elections. Artificial intelligence has made it possible for almost anyone to create convincing fake videos, cloned voices, fabricated documents and manipulated images. The technology is no longer locked away in film studios or limited to cybercriminals with special skills. Free and cheap AI tools can copy someone’s face, mimic a public figure’s voice, and produce official-looking paperwork in minutes. Election experts warn that this synthetic content could confuse voters and seriously undermine confidence in the electoral process. If people cannot trust what they see and hear, they may stop trusting the election entirely. The Independent National Electoral Commission, INEC, has acknowledged that this is a growing threat. It says it has monitoring mechanisms, an AI division and a rapid-response team to detect and respond to fabricated content. It has also urged Nigerians to rely only on information and results released through its verified communication channels. But as the report shows, the scale of the problem may be bigger than any single institution’s ability to handle it, especially when the 2023 elections already offered a painful lesson in how fast false information can spread.
INEC’s response is not passive, but it is under pressure. Wilfred Ifogah, Deputy Director of Voter Education and Publicity at INEC, said the commission has constantly maintained that Nigerians should rely only on results from INEC through its official communication channels. That is the first line of defence, but not the only one. Ifogah disclosed that the commission has its own monitoring mechanism and an AI division, and that it will constantly put disclaimers on fake content once detected. He also revealed the existence of a dedicated rapid-response team and a vibrant media monitoring unit, which respond as soon as false content is flagged. More importantly, INEC is working with major technology and social media platforms including Meta, X, TikTok, Google and WhatsApp to tackle election-related misinformation and disinformation. On paper, this sounds like a serious effort to protect the credibility of electoral activities. But there are real questions about speed and practical reach. During the 2023 general elections, false claims, manipulated materials and misleading election content circulated on a massive scale. Some were designed to look like official INEC communication. Others were old videos pulled out of context and presented as current. The repeated message from fact-checkers was that misinformation was not just a nuisance; it was a direct threat to democratic integrity. Now that AI has arrived, the same technology used by platforms to moderate content is also being used by bad actors to create more convincing fakes. The defenders and the attackers are using the same tools. The difference is that attackers only need to be right once, while defenders have to be right every time.
The 2023 election period was a turning point for information integrity in Nigeria. The Nigerian Fact-Checkers’ Coalition, known as NFC, responded by setting up election situation rooms in Abuja, Lagos and the United Kingdom to monitor misinformation and disinformation capable of inciting violence, promoting voter apathy, undermining trust in the electoral process and influencing the outcome of the elections. Opeyemi Kehinde, editor at FactCheckHub and coordinator of the NFC, said at the time that coalition members would monitor and debunk false information from election day until INEC announced final results. He understood that misinformation could never be allowed to have its way in the electoral process. The numbers from that exercise are sobering. During the presidential and National Assembly elections, and the governorship and State Assembly elections, fact-checkers verified 127 claims. Videos accounted for a significant proportion of the materials subjected to verification, which made the work even harder. Fabricated media graphics and old videos presented as though they were connected to the current elections also circulated widely. What made the situation worse was that many of these false items were shared by ordinary people who believed they were helping, not harming. They forwarded a video because it seemed important, not because they wanted to deceive. But the effect was the same: false information spread quickly, and by the time corrections were released, the false story had already settled in people’s minds. For anyone who lived through that period, the lesson was unforgettable. Misinformation is not a minor problem that can be cleaned up after the election. It is a force that can shape public perception and even influence the outcome. And in 2023, the tools available to create fake content were far less powerful than the ones available now.
The new danger is generative AI. Abuja-based cybersecurity and AI expert Salami Daniel put it plainly: artificial intelligence has significantly reduced the barrier to producing convincing fake audio, videos, images and documents. This is not a small improvement; it is a transformation. In the past, creating a convincing deepfake required technical expertise, expensive software, high-quality computing power and a lot of time. Today, a single AI tool can clone a person’s voice from a short recording, imitate their facial expressions from publicly available videos, and generate documents with the look and feel of official electoral materials. Daniel says this creates a serious verification challenge ahead of the 2027 elections. A fabricated video or audio recording can be produced within minutes and circulated widely before the affected person or institution has an opportunity to respond. The danger is especially acute for INEC. Imagine a fabricated video attributed to the commission, showing an official announcing that results from two states have been cancelled. Even if INEC quickly says the video is fake, tens of thousands of people may have already seen it. Some will never see the correction at all, because WhatsApp forwards spread inside closed groups that are difficult to monitor. Others, for political or personal reasons, will choose to believe the fake even after it has been debunked. Daniel therefore argues that the response cannot stop at debunking false content after it has gone viral. There must be mechanisms to rapidly authenticate official results, statements, videos and other electoral materials. There must be continuous monitoring of social media platforms, and there must be better coordination between INEC, security agencies, civil society and the technology platforms themselves. If people have a fast and reliable way to check whether something is true, the power of the fake is greatly reduced.
Responsibility also falls on journalists and ordinary citizens. Opeyemi Kehinde has urged journalists to exercise caution before publishing or amplifying unusual election-related audio, videos, screenshots or documents. He says the material should be verified from the original source and cross-checked with the relevant authority before being shared with the public. In an online environment that rewards speed, this is not always easy. A reporter who waits too long to verify a story might be beaten by another outlet that only cares about clicks. But the cost of amplifying a dangerous falsehood is far greater than the cost of being a few minutes late. Kehinde also warns that no single detection tool is completely reliable. AI can sometimes spot AI-generated content, but not always. As the technology improves, detection tools may struggle to keep up. The most effective approach, he says, is likely to combine technical detection, human verification, source authentication and rapid communication from trusted institutions. Another fact-checker, Kamaru Danlami, who spoke to the correspondent in Kaduna, agrees. He says the 2023 elections demonstrated how quickly misinformation and disinformation could spread during voting, result collation and the announcement of results. AI now gives those producing false information more sophisticated tools to make fabricated content appear authentic. A fake result, a manipulated video, or a cloned voice attributed to an electoral official could circulate widely before fact-checkers, journalists or the affected institution have established that it is false. Danlami also notes that the danger extends beyond fabricated election results to false claims about voting procedures, polling units, candidates, electoral officials and incidents at polling stations. All of this can create fear, anger, confusion and distrust. That is why media literacy is so important. Voters must learn to question the source of election-related information, especially when it is sensational and has no clear origin. At the same time, INEC, political parties, the media, technology platforms and fact-checking organisations need to build a much stronger coordination system so that false information can be identified, verified and countered quickly, with one clear voice.
The bottom line is that the challenge before INEC is not just about detecting fake results, videos and audio. It is also about making sure verified information reaches voters quickly enough to prevent false narratives from gaining ground. The commission says its AI division, rapid-response team and media-monitoring mechanisms are deployed for this purpose. It continues to urge Nigerians to verify election-related information through its official channels. But the report makes clear that this will not be enough if the rest of the country is not involved. The 2027 elections will likely be fought not only at polling units but also in group chats, on social media, and through the constant flow of audio, video and documents into ordinary people’s phones. If the information space is poisoned, even a well-run election can lose public credibility. The ability of INEC, technology platforms, fact-checkers and the media to promptly authenticate and communicate electoral information will be central to maintaining public confidence. But ordinary citizens also have a role to play. They can pause before forwarding a shocking video. They can ask where a piece of information came from. They can seek out official sources before making up their minds. It may sound like a small action, but in a society where a single viral clip can ignite suspicion or violence, individual caution is a form of civic duty. The experts in the report shared one lesson from 2023: information can become part of the electoral contest itself. With generative AI becoming more accessible, fabricated content will only become more sophisticated and harder for ordinary users to identify. This means the defenders of electoral truth need to be just as creative and fast as the attackers. It is a tough race, but not impossible. The truth may not always travel first, but it can still win if enough people are committed to helping it move faster.