By Lemmy Ughegbe, Ph.D
For generations, human beings have relied upon a simple rule of evidence: seeing is believing. Artificial intelligence is rapidly destroying that certainty.
By Nigeria’s 2027 general election, a voter may watch a presidential candidate announce his withdrawal from an election he has not withdrawn from.
She may hear another candidate insult an ethnic or religious group in words he never uttered.
Millions may watch a politician apparently admit to corruption, electoral manipulation, or some other scandal when, in reality, the person neither spoke those words nor appeared before the camera.
His face might be the one. The voice may sound exactly like hers. The event may never have happened. That is the frightening electoral frontier Nigeria is approaching.
The Independent National Electoral Commission has warned that artificial intelligence-driven disinformation poses a serious threat to the credibility of the 2027 elections.
INEC Chairman, Professor Joash Amupitan (SAN), made the point last week at a communication scholars’ conference in Abuja, warning that generative AI has made manipulated video, audio, and images increasingly cheap to produce and difficult to detect.
He is right to be worried.
Nigeria has spent decades trying to ensure that votes are genuine. In 2027, we may also have to ensure that the information upon which people cast those votes is genuine. That changes the nature of election management.
Until relatively recently, political misinformation required considerable effort. Readers could challenge a false newspaper story. A manipulated photograph often left visible traces. People could sometimes dismiss an impersonated voice as obviously fake.
Some synthetic audio and video are already convincing enough to mislead ordinary viewers, especially when released at high-pressure moments. Detection is possible but often slower than the extraordinary viral speed and spread.
Social media then supplies the distribution network. WhatsApp, Facebook, TikTok, X and Instagram can move fabricated content from one telephone to millions before journalists, fact-checkers or electoral authorities have established the truth.
The danger is acute during elections because political falsehoods do not need to survive forever. Sometimes they only need to survive long enough.
Imagine a convincing video appearing at 6 am on election morning in which a leading candidate supposedly tells supporters to boycott the election because INEC has compromised the process. Or an audio recording apparently revealing that security agencies have been instructed to favour a political party.
If millions believe the fabrication for five hours, correcting it five days later becomes almost meaningless. The damage would already have been done.
This is why INEC’s proposed rapid response mechanism, which the Commission says will counter falsehoods within the same news cycle, matters. But INEC cannot fight this battle alone.
The media must recognise that speed, which has become journalism’s obsession in the digital age, can become democracy’s enemy when verification disappears.
A newsroom that rushes to publish sensational audio because competitors already have it may unknowingly become the distribution arm of political disinformation.
The old journalistic discipline of verification therefore becomes even more valuable in the age of artificial intelligence.
Who produced this material? Where did it originate? Can the person supposedly speaking confirm it? Can we examine its metadata? Has an independent party authenticated it?
These questions may slow publication by minutes. Those minutes may protect democracy.
Political parties also carry responsibility. It would be hypocritical for politicians to complain about fake news while maintaining propaganda operations whose purpose is to manufacture and distribute falsehoods against opponents.
There must be consequences for deliberate digital deception. Our legal and electoral systems must respond if campaign organisations knowingly create fraudulent material to manipulate an election.
Technology companies cannot escape responsibility either. Platforms that profit from political engagement must invest in detecting manipulated content, identifying coordinated disinformation campaigns, and responding rapidly when electoral fabrications spread.
But perhaps the most important defence is the Nigerian voter. INEC has correctly identified media literacy as part of the solution. Citizens must acquire a new democratic reflex: do not immediately believe everything you see, hear, or receive.
That may sound simple. It is not. Political disinformation succeeds because it frequently confirms what people already want to believe.
A supporter readily believes scandals about an opponent. An ethnic grievance makes an inflammatory fabrication seem plausible. Religious suspicion lowers the threshold of scepticism. Partisanship often defeats verification before verification even begins.
Artificial intelligence therefore does not create all our vulnerabilities. It weaponises existing ones. Algorithmic micro-targeting should concern us just as much as deepfakes. Different political realities can be shown to different groups.
A candidate might seem moderate to one group but extreme to another. While some messages leverage religion to reach religious communities, others use ethnic narratives to connect with different audiences.
There may no longer be one national political conversation. There could be millions of personalised ones. That threatens the shared factual foundation upon which democratic debate depends.
Nigeria already enters elections burdened by distrust. Some citizens distrust INEC. Political parties distrust one another. Government and opposition routinely accuse each other of manipulation. Sections of the public distrust traditional media.
Into this fragile environment comes technology capable of manufacturing extraordinarily convincing evidence. The implications should worry everyone.
INEC’s response must therefore go beyond issuing occasional disclaimers. There should be an easily identifiable, continuously staffed election information centre capable of authenticating or debunking viral claims quickly.
News organisations should establish dedicated verification desks. Political parties should agree to protocols against synthetic electoral deception. Civil society organisations should intensify digital literacy campaigns.
And Nigerians should learn perhaps the most important electoral lesson of the AI age: pause before you share, verify before you believe.
The ballot remains the physical instrument of democracy. But increasingly, people will fight the battle for that ballot on screens long before voters arrive at polling units.
Nigeria must prepare accordingly. Because the greatest threat to the credibility of the 2027 election may not be a fake ballot. The greatest threat could be a fake reality, which is convincing enough to determine how a genuine ballot is cast. In 2027, seeing may no longer be believing. Verification must become the new seeing.
Lemmy Ughegbe, FIMC, CMC
Email: lemmyughegbeofficial@gmail.com
WhatsApp ONLY: +2348069716645
