On a Monday in Abuja, as Nigeria looks toward the next general elections, a warning was issued not about ballot boxes, but about the stories told around them. At the National Summit on AI, FIMI and Electoral Information Integrity, the Ambassador of the European Union to Nigeria and ECOWAS, Mr Gautier Mignot, offered a sobering thought: elections can be undermined without a single ballot being altered. Fabricated announcements, false claims about results, and carefully timed lies can do what stolen ballot boxes once did. The immediate target of such manipulation, he said, is information; the deeper target is trust. Disinformation does not need to convince everyone. It only needs to create enough confusion to make people wonder whether the process is real, whether their vote matters, and whether the official results can be believed. Mignot urged Nigeria to prepare early, far ahead of 2027, rather than waiting for the storm to arrive. He reminded his audience that this is not a uniquely Nigerian problem or an African problem. Many democracies, including countries in Europe, have faced foreign interference and manipulation of information during elections. It is a global phenomenon, and it has learned to travel quickly across borders. In that sense, Nigeria is not being singled out; it is being invited to learn from the mistakes and experiences of others before the next election cycle becomes a battlefield of narratives.
Artificial intelligence, Mignot said, is not simply an enemy of democracy. It can support translation, improve accessibility, and open new channels of communication. But the same technologies can make deception easier and harder to recognize. A voice can be cloned. A face can be animated. A completely fictional event can be made to look real. The task, he insisted, is broader than detecting deepfakes. It is about building a civic immune system that can resist manipulation before it spreads. He recalled lessons from the European Union’s 2023 Election Observation Mission, which highlighted the importance of timely public information, independent fact-checking, and the capacity to address harmful content in local languages. These are not abstract recommendations. They are practical tools that can determine whether a rumour becomes a crisis or remains just a rumour. Mignot offered three priorities for 2027. The first is preparation before pressure. Verification contacts must be established now. Institutions must clarify who authentically communicates official messages and who people should turn to when false directives are circulating. Confusion during an election is dangerous because it supplies the raw material for conspiracy theories. If citizens already know which channels to trust and which voices are official, a fabricated directive can be dismantled quickly rather than allowed to spread for days. The second priority is making partnerships operational. International cooperation must move beyond meetings and goodwill gestures. It should mean coordinated responses, shared analysis, and technical support, while respecting that each institution has its own mandate. Coordination does not mean fusion; it means knowing who does what and ensuring that no gap is left open. The third priority is the most delicate: protecting democratic rights while protecting information. Countering manipulation must never become an excuse to silence criticism. A free society cannot defeat disinformation by becoming less free. Mignot stressed that the answer is not censorship but transparency, verification, and an empowered public.
He also drew attention to those who are often hit hardest by digital abuse. Women, especially women in public life, are frequent targets of impersonation, harassment, and violent online abuse. Elections cannot be fair if some voices are frightened into silence. Disinformation often thrives by exploiting inequalities, and any serious strategy must include protections for women and for groups that are already marginalized. It must also ensure that information is accessible across all languages and to persons with disabilities. The fight for information integrity is not only about stopping falsehoods; it is about making sure that every citizen, regardless of language, literacy, or disability, can find and understand the information they need to participate meaningfully. Mignot was careful to frame the international role correctly. Nigeria’s electoral choices belong to Nigerians, he said. The European Union is not there to decide outcomes or to impose solutions. Its role is to support Nigerian-led efforts, to share what European democracies have learned, and to stand beside those working to strengthen the integrity of the electoral process. That message was central to the summit’s spirit: external actors can help, but the responsibility for building trust rests at home. His warning was the opening frame; the Independent National Electoral Commission immediately filled in the institutional response.
INEC Chairman, Prof. Joash Amupitan, announced that the commission plans to establish an Artificial Intelligence unit to tackle misinformation, disinformation, and deepfake content. He described a dramatic shift in electoral vulnerability. Today, he said, elections can be undermined long before the first ballot is cast, through synthetic media, automated misinformation, deepfake fabrication, digital micro-targeting, and foreign information manipulation. Hyper-realistic fake videos of officials can be created at negligible cost. Bot networks can simulate outrage, manufacture trends, and make a false story appear to be a groundswell of public opinion. Amupitan identified specific windows of vulnerability: pre-election operations, the campaign period, and the collation of results. During Continuous Voter Registration, for example, forged registration portals could trick citizens into submitting their details to the wrong hands. Fake operational directives could confuse officials and disrupt preparations. He cited an example that should make every election manager pause: during the 2023 Slovak parliamentary election, an AI-cloned audio recording disrupted the political landscape during a blackout period, when voters were supposed to have time to reflect. Similar tactics could be used to distribute synthetic audio falsely announcing results, creating panic and undermining confidence in the official process. The deep concern is not only with the obvious deepfake that is eventually exposed, but with the more subtle fabrication that arrives at exactly the wrong moment. Amupitan argued that INEC cannot ignore these dangers because modern electoral technologies already depend on artificial intelligence. The smart card reader and the Bimodal Voter Accreditation System already have AI tools embedded. The same technology that makes elections more efficient can also be turned against them. His warning was blunt: any country, any institution, that folds its arms and does not begin to address the dangers posed by AI is planning to fail. The decision to create an AI unit is not a luxury; it is an attempt to keep up with the threats.
But Amupitan acknowledged that INEC cannot police truth alone. No election management body can monitor every post, verify every video, and correct every false claim by itself. He called for fast-track escalation pathways with platform operators, clear communication channels with telecoms and media, and direct access for fact-checkers who can verify claims in real time. He also proposed simulations of AI-driven crisis scenarios ahead of the 2027 elections, so that the response is not invented under pressure but practiced in advance. This kind of preparedness is not about eliminating false information completely; that is impossible. It is about making sure that when false information appears, the right institutions, the right people, and the right messages are ready to respond quickly and clearly. The Minister of Information and National Orientation, Mohammed Idris, represented by Diri D. Ezugo, added another essential dimension: the media must remain free to question government, political parties, candidates, and public institutions. Critical journalism must not be mistaken for unfavourable journalism. In a healthy democracy, the press is not an adversary of the state; it is a guardian of accountability. But the information environment also shapes electoral integrity. Minister Idris noted that integrity extends far beyond voting and collation. It depends on the quality of information that citizens receive before, during, and after the process. He cited Nigeria’s National Artificial Intelligence Strategy 2024 and the establishment of the UNESCO Category 2 International Media and Information Literacy Institute as evidence of the country’s commitment to credible information. He also referenced the Electoral Act 2026, signed on February 18, 2026, which provides for mandatory electronic transmission of results and recognizes the IReV portal. These are important institutional steps. But technology alone, he cautioned, cannot guarantee confidence. The response to misinformation must remain lawful and rights-respecting. If the cure for disinformation becomes arbitrary censorship or the drowning out of legitimate dissent, then the election may be technically smooth but politically hollow. The goal is not to silence sounds but to preserve the signal.
Former INEC Chairman, Prof. Attahiru Jega, who chaired the summit, offered a practical suggestion that connected the concerns of the day with a manageable, citizen-friendly solution. He urged INEC to deploy chatbots with officially verified data to provide accurate electoral information. He recalled that the 2024 Cambridge Conference on Electoral Democracy emphasized the need for African Election Management Bodies to use chatbots so that citizens can quickly obtain and verify information. Some election management bodies have already found that these tools reduce misinformation because they give people a direct, trusted source to check claims. Instead of asking citizens to navigate a chaotic media landscape, a well-built chatbot can answer questions about voter registration, polling units, result processes, and official announcements. It can also act as a first line of defense against lies, offering quick access to the truth. Jega’s suggestion is grounded in a simple insight: people want reliable information, but they often do not know where to find it. The official page may exist, the press release may be written, but if a citizen on a busy afternoon receives a false message on WhatsApp, they need an immediate, accessible way to check it. A chatbot can be that response. The summit’s message, taken as a whole, was not anti-technology or anti-AI. It was a call for deliberate preparation, for stronger institutions, for honest journalism, and for citizens who are equipped to ask questions. The ballot remains secret, but the information environment is public. Nigeria’s electoral choices belong to Nigerians, and the responsibility for protecting them is ultimately Nigerian. External partners, from the European Union to international observers, can offer lessons and support. They can share research, fund fact-checking, and help build capacity. But they cannot manufacture trust. Trust is built in the months and years before election day, in the decisions made by electoral officials, in the integrity of the media, in the behavior of political leaders, and in the daily experience of ordinary citizens who see whether the system works. As 2027 approaches, the challenge is not to predict every false story or block every fake video. The challenge is to create an environment where truth has a fighting chance, where citizens are not left alone against an army of algorithms, and where the deepest target of disinformation, trust, is protected before the pressure arrives.