As Nigeria moves closer to the 2027 general elections, artificial intelligence is creating a new challenge for voters: the rapid spread of convincing fake videos, audio recordings and images capable of making people appear to say or do things that never happened.
The concern is already being raised by election authorities as the Chairman of the Independent National Electoral Commission, Joash Amupitan, recently warned that AI-driven disinformation poses a serious threat to the credibility of the 2027 elections, noting that generative AI has made manipulated audio, video and images cheaper and increasingly difficult to detect.
How AI Can Manipulate Political Content
A politician could be made to appear as though they made an inflammatory statement they never uttered. Their voice could also be cloned and used to create a fake speech, phone call or WhatsApp voice note.
But the danger is not limited to completely fabricated content.
Real videos and audio recordings can be edited, shortened or taken out of context. An old speech can be presented as a recent statement, while fake newspaper front pages, campaign posters and screenshots can be designed to look authentic.
This makes the problem particularly difficult because voters may not always be dealing with completely fake material. A genuine recording can be manipulated in a way that changes its meaning.
Social Media Could Amplify the Problem
With social media and messaging platforms playing an increasingly important role in how Nigerians consume news, false political content can spread rapidly before journalists, fact-checkers or authorities have time to verify it.
A recent SBM Intelligence survey found that 66.5 percent of Nigerians surveyed were extremely or very concerned about AI-generated political misinformation, while WhatsApp was identified as the primary news source for 52.1 percent of respondents.
The problem could become even more complex when AI-generated content is produced in local languages. Experts have warned that cloned voices and synthetic messages could impersonate politicians, traditional rulers, religious leaders or community figures and circulate through WhatsApp and other platforms.
The Danger of the “Liar’s Dividend”
There is another threat: genuine recordings could also be dismissed as AI-generated.
As deepfakes become more common, politicians or supporters accused of making controversial statements could simply claim that authentic recordings are fake. This can create what researchers call the “liar’s dividend”, a situation where the existence of widespread fake content makes people less willing to believe genuine evidence.
That could weaken public trust even when a recording is authentic.
A fake video of a candidate making an offensive statement, for example, could go viral before it is debunked. By the time the correction arrives, thousands or even millions of people may already have seen it and formed an opinion.
What Voters Can Do
Nigerians do not need to become technology experts to reduce their exposure to AI-generated misinformation.
Before sharing political content, voters should ask:
- Who originally posted it?
- When was it recorded?
- Where is the original video or audio?
- Has a credible news organisation reported the same claim?
- Has a trusted fact-checking organisation verified it?
- Is the content being shared without context?
A familiar face or recognisable voice should no longer be treated as proof that a recording is genuine.
Voters should also be cautious about emotionally charged content designed to provoke anger, fear or outrage. Such material is more likely to be shared quickly without verification.
Protecting the 2027 Election
The threat posed by AI means election preparedness will have to extend beyond polling units, ballot papers and electoral technology. It will also involve protecting the information environment in which voters make their decisions.
INEC has acknowledged the growing threat and is conducting an audit of its electoral technology systems as part of preparations for the 2027 polls.
Political parties, journalists, technology companies, fact-checkers and civil society organisations will also have important roles to play in identifying and correcting false information quickly.
Ultimately, voters remain the first line of defence.
In an election increasingly shaped by artificial intelligence, the most important habit may be a simple one: stop, check, cross-check and then share.
