Nigeria's upcoming election marred by rising AI disinformation

Nigeria has just over two months to go until its general election, with incumbent Bola Tinubu running on a ticket focused on combating raging insecurity and resolving the nation's economic crisis. He faces off against former Vice President Atiku Abubakar and senior opposition figure Peter Obi.

The country's electoral body, the Independent National Electoral Commission (INEC), recently released its preliminary voter's register, which reportedly is riddled with images believed to have been generated using artificial intelligence (AI).

Among the telltale signs of AI usage were exaggerated facial features in images and inconsistencies on the electoral registers, such as duplicates of names and even identities of people who since the previous election had died.

Is Nigeria becoming a one-party state?

The development raises concerns about the body's overall preparedness ahead of the country's upcoming 2027 general election. Instead of making the upcoming vote more transparent by allowing access to its procedures, INEC has come under fire amid broader controversies about the expanding influence of AI in elections across the continent.

Fake news: Nothing new in Nigeria

Joash Amupitan, the chairman of the electoral commission, said the body is now investigating the development, focusing on discrepancies potentially created by AI use.

To this end, he announced the establishment of a dedicated unit to monitor AI-related threats, detect deepfakes and strengthen the integrity of the electoral process, warning that the credibility of the upcoming election could otherwise be under attack.

However, even prior to the proliferation of AI across Africa and the globe, this would not be the first time that Nigeria has found itself in the unfortunate position of having to deal with the circulation of fabricated images, doctored audio and fake videos online. In the run-up to the 2023 election, deepfakes falsely linking candidates to terror-related activities filled social media platforms.

These are often designed to promote certain candidates or to discredit opponents in a bid to influence public perception, which now with the use of AI could create disinformation campaigns of new proportions.

Artificial intelligence has various effects on voter behavior

Lois Ugbede, the assistant editor at the factchecking platform DUBAWA, told DW that "based on our monitoring so far, AI-driven disinformation is serious and is becoming commonplace."

"This could influence political discourse and voter behavior ahead of 2027 because we are already seeing parties using inflated crowds to show that they have a large followership or support," she added.

Have some elections become an empty ritual?

At the same time, rather than influencing the vote in favor of any particular candidate, "the deeper danger is that people may begin to doubt authentic information simply because they know convincing AI-generated fakes are possible," said Akintunde Babatunde, executive director of the Abuja-based Centre for Journalism, Innovation and Development.

"We often focus on the risk of people believing an AI-generated falsehood, but there is another problem where genuine evidence can be dismissed as AI-generated," he noted, citing instances of a lawmaker who recently made inflammatory remarks during a campaign rally — but then claimed the clips were manipulated.

Further pitfalls behind a lack of safeguards against AI

Chibuike Mgbeahuruike, an election consultant at the Centre for Democracy and Development, told DW that apart from the spread of synthetic content online, AI intrusion into the electoral commission's portal could also "undermine its institutional credibility, especially with the low level of confidence that we are already witnessing."

He added, however, that AI itself is not inherently a threat to electoral integrity. The real threat lies in "the absence of effective accountability for its deployment," as innovations in tech increasingly appear to infiltrate governments and their democratic processes.

"Voters must have confidence … that the technology is also transparent, accountable and independently scrutinized, and ultimately subject to human and institutional oversight," said Mgbeahuruike.

Mallam Jibrin Baba Ndace, the director-general of state-owned broadcaster Voice of Nigeria, agrees, urging journalists at a recent conference also to not fall into the trap of believing reports, which later could turn out to be false due to AI manipulation.

He stressed the importance of protection measures and due diligence in news media, with the growing inability of public institutions like INEC to provide impartial information due to the use and influence of AI and other election-related tech.

Nigeria's marred relationship with new tech

Beyond AI, Nigeria's electoral processes have undergone various technological shifts over the years in hopes of improving transparency while increasing efficiency.

Electoral authorities first deployed electronic assistance in 2011, when Direct Data Capture Machines were introduced; the devices collect voters' biometrics by scanning fingerprints or capturing facial recognition data with webcams to stop voters from submitting multiple ballots.

Ten years later, a system called the Bimodal Voter Accreditation System (BVAS) was introduced as the new standard, designed to combine fingerprints with facial recognition. However, during the 2023 general election, this voting process was marred by system failures, as poor cell signals delayed these voter accreditation procedures.

However, with INEC tallying more than 103 million Nigerians registered to participate in the coming poll, disinformation, fake news and technological challenges amount to only a few of several concerns pertaining to Nigeria's democracy.

Voter apathy has become a particular concern, especially among the country's youth. The 2023 election recorded the lowest voter turnout since the country's return to democratic rule in 1999, with merely 27% of registered voters going to the polls.