Beneath the Surface By Dakuku Peterside
Beneath the Surface By Dakuku Peterside
A voter who has spent hours in a queue should not need to understand computer servers to believe that a ballot will be counted. The obligation runs the other way. Those entrusted with conducting an election must demonstrate that its outcome reflects the choices citizens made. As Nigeria approaches 2027, this should be the measure of every technological innovation: does it make electoral power more accountable to the people whose consent gives that power legitimacy?
The attraction of technology is understandable. Where human discretion has attracted suspicion, a machine can seem to promise impartiality. Biometric checks offer firmer identification; digital records offer opportunities for scrutiny. Artificial intelligence could improve logistics and detect inconsistencies. Yet the promise becomes dangerous when people mistake sophistication for integrity. Every system still reflects decisions about procurement, access, oversight and what happens when something goes wrong. Those decisions remain human, even when their consequences appear on a screen.
Nigeria’s 2023 elections exposed the fragility of the relationship between technical capability and public confidence. The Bimodal Voter Accreditation System, BVAS, sought to strengthen voter identification through biometric checks. The INEC Results Viewing Portal, IReV, enabled public access to images of polling unit result sheets. Their functions differed, but their democratic value lay in making key stages of the process more open to verification.
Delays in uploading presidential results undermined that promise. They did not, by themselves, establish that votes had been altered. Equally, the lack of expected information mattered. Citizens were entitled to ask what failed, what remained secure, and how to check the records. A transparency mechanism cannot build confidence through its promised availability and then become irrelevant when people need to consult it. Its failure requires an explanation proportionate to the trust invested in it.
Behind these questions stood people who had paid a price to participate. A trader who closed a stall had surrendered income. A worker who travelled home to vote had committed money and time. When the count became difficult to follow, their uncertainty concerned more than a preferred candidate. It touched the belief that peaceful participation mattered. An election can leave institutions formally intact while making citizens less willing to place their hopes in them again.
This is the democratic cost Nigeria must avoid in 2027. Confidence cannot be repaired by asking people to ask fewer questions. It grows when institutions give better answers and make those answers examinable. Before deploying AI, INEC should explain its intended tasks, the data it will process and the suppliers involved. Responsibility for errors must be identifiable in advance. Public scrutiny belongs before deployment, where it can still influence decisions, and after failure.
Communication during disruption deserves its own preparation. An institution should be able to say what it knows, what remains uncertain, and when it will provide another update. Silence invites speculation; premature certainty can destroy credibility when later evidence contradicts it. Clear communication also protects officials from having every operational problem interpreted as deliberate manipulation. It gives the public a basis for distinguishing an interruption from a breach and a documented correction from an unexplained alteration. The aim is neither automatic suspicion nor automatic acceptance, but judgement informed by records that remain available for independent examination after the immediate political controversy has passed.
Independent testing should therefore produce evidence of readiness under actual operating conditions. The examination must include interrupted power, poor connectivity, heavy demand, and uneven technical support. Publish relevant findings and corrective actions without exposing sensitive information. A successful demonstration is reassuring; a serious attempt to discover failure is more useful. The public needs to know which weaknesses were found, which were corrected and which limitations still require practical safeguards.
Those safeguards must reach the polling unit. Officials need clear, lawful instructions for equipment and network problems, supported by rehearsals and accessible technical assistance. Voters and observers should understand the applicable procedures. Otherwise, discretion exercised under pressure can appear indistinguishable from favouritism. Preparedness also means designing for elderly voters, people with disabilities and people whose biometric details are difficult to capture. Efficiency has little democratic value if its apparent success conceals citizens who could not participate.
Verification must continue through the count. A traceable link must exist between the polling unit record, its published image, and the figures carried through collation. Where discrepancies appear, a functioning process must investigate and explain them. Merely allowing citizens to see a problem is insufficient. Visibility without a route to correction can deepen the sense of powerlessness. Technology should make irregularities harder to conceal and evidence-based challenges easier to pursue.
AI can support this work, but its authority needs firm boundaries. A system that flags a suspicious record raises a question; it does not establish wrongdoing. Decisions affecting registration, eligibility or results must remain subject to authorised human judgement and established procedures. Human oversight must involve someone who can examine the evidence, challenge the output, and explain the decision. A signature beneath an opaque recommendation offers weak protection if the signatory cannot meaningfully assess its basis.
The information environment presents a different danger. A cloned voice could announce a false withdrawal; a fabricated security warning could discourage turnout. Such deception can damage an election without touching a result sheet. Worse, convincing fabrications create an excuse to dismiss authentic evidence as fake. Citizens may increasingly believe whatever confirms their loyalties. When disagreement extends to whether any reliable account of events is possible, democratic argument loses the common ground on which persuasion and accountability depend.
This danger becomes intimate when misinformation arrives through trusted relationships. A relative’s forwarded message carries social authority that its contents may not deserve. Citizens should pause before sharing material designed to provoke fear or fury. But individual caution needs institutional support. Electoral authorities, journalists, platforms and civil society should establish coordinated arrangements for early warning, verification and rapid correction. Reliable information must reach people through radio and in Nigerian languages, with clear channels to check urgent claims.
Political parties must answer for the persuasion they organise. Standards against deceptive impersonation should govern campaign staff, contractors and commissioned content. Candidates should correct false claims produced by their campaigns and repudiate dangerous fabrications circulated in their name. Demanding integrity from electoral officials while rewarding dishonesty among supporters hollows out the principle being defended. The duty to respect voters begins with how their choices are influenced, long before those choices are counted.
Journalists face a related discipline. Reporting that a platform failed differs from explaining why it failed or what happened to the underlying records. Speed should not erase that distinction. Observers and fact-checkers need preparation and access sufficiently to examine evidence. Platforms must strengthen reporting channels and responses to deceptive synthetic content in Nigerian languages. Verification becomes more credible when its methods are transparent, and its standards apply consistently across political interests.
Cybersecurity requires continuous work: controlled access, protected backups and rehearsed recovery plans. Yet security cannot be a universal answer to accountability demands. Nor should combating disinformation justify suppressing legitimate criticism. Citizens must remain free to question electoral authorities, including sharply. The institutional response should distinguish demonstrable deception from unwelcome scrutiny. Protecting the election means preserving the freedom to expose its weaknesses and demand correction.
Suppliers, too, must accept obligations that survive the sales presentation. Electoral systems need dependable support, auditable records and clear contractual responsibility for failures. Commercial confidentiality should not make essential questions unanswerable. Developing African technical expertise and regional cooperation could strengthen independent assessment and reduce reliance on suppliers’ assurances. The lasting investment is institutional knowledge: people who can examine systems, investigate incidents, and retain expertise after a contract ends.
Shared responsibility requires careful language. Voters cannot repair infrastructure they did not procure or disclose records they do not control. Accountability must follow authority. INEC answers for administration, suppliers for their obligations and parties for campaign conduct. Each actor needs duties against which performance can be judged. Otherwise, collective responsibility becomes a comfortable way to distribute concern while detaching the power to prevent failure from the consequences of failing.
Citizens also face a test of consistency. Do we require the same evidence when our candidate wins as when our candidate loses? Excusing an irregularity because it benefits our side weakens a protection we may need tomorrow. Standards that change with the winner cannot sustain trust. Credible elections require the willingness to defend a fair process even when its outcome disappoints, while retaining the right to challenge wrongdoing through evidence and lawful procedures.
Before 2027, Nigeria should demand measurable preparedness: published governance rules, independent audits, trained personnel, tested contingency plans and functioning complaints mechanisms. Each commitment needs a responsible institution, a deadline and public evidence of completion. The question is whether citizens will encounter a process they can understand, scrutinise and challenge. They have already supplied their patience and their choices. Those entrusted with conducting the election must now supply the competence and proof that make those choices count.
•Dr Dakuku Peterside is a turnaround expert in leadership and management.