morganable Politics/elections
The commission also revealed that some genuine photographs submitted by voters had been digitally enhanced using AI-powered applications on mobile phones.
KaNo —
As preparations intensify for the 2027 general elections, artificial intelligence experts have urged the Independent National Electoral Commission to deploy advanced AI-powered tools to identify irregularities in the voters’ register, detect duplicate registrations and strengthen the credibility of the electoral process.
The experts, who spoke during an interview with Punch ahead of the forthcoming PUNCH Media Foundation webinar on Artificial Intelligence, said the technology, if properly deployed, could improve electoral transparency, enhance voter education and help the commission identify potential threats before and during elections.
They, however, warned that the growing accessibility of generative AI tools could also expose Nigeria’s electoral process to manipulation through deepfake videos, fabricated audio recordings, misleading photographs and false announcements capable of discouraging eligible voters from participating in elections.
One of the concerns, they said, was the possibility of malicious actors using AI-generated notices to falsely announce the postponement or cancellation of voting in particular polling units or communities, thereby preventing voters from exercising their franchise.
The warning comes amid growing concerns over the use of artificial intelligence to manipulate election-related information ahead of the 2027 polls.
The concerns were heightened following the circulation of AI-generated images during the June 20, 2026 governorship election in Ekiti State.
During the election, an image purporting to show voters queuing to sell their votes circulated widely on social media.
However, fact-checkers at the Centre for Democracy and Development established that the image was generated using artificial intelligence.
Another image circulated online as evidence of massive voter turnout in Ado-Ekiti, but an analysis by the Nigerian Fact-checkers’ Coalition also found that it was AI-generated.
Beyond the circulation of misleading election content, INEC has also uncovered suspected AI-generated photographs and other irregularities in its voters’ register.
Last week, the commission disclosed that its biometric system had flagged 268,000 records over suspected AI-generated photographs and other irregularities.
INEC further acknowledged that some of the questionable records were introduced into the register by internal and field registration officers who allegedly used their authorised login credentials to manipulate the registration process.
The commission also revealed that some genuine photographs submitted by voters had been digitally enhanced using AI-powered applications on mobile phones.
Providing further details, INEC’s Director of Information and Communication Technology, Dr Lawrence Bayode, told Punch that the system had flagged 268,000 records but declined to disclose the number of records that had so far been reviewed.
Bayode explained that the Automated Biometric Identification System was designed to identify multiple registrations, underage voters and other irregularities in the registration database.
He said the commission had commenced a human adjudication exercise to examine the flagged records and remove those found to be fraudulent.
“Human Adjudication exercise will remove them. Those that we have seen, we are cleaning up,” he said.
The ICT director, however, clarified that voters whose genuine photographs had been enhanced using AI applications would not automatically be excluded from the register.
According to him, there is a distinction between photographs originally taken of genuine voters and subsequently improved using AI-powered applications and images generated entirely by artificial intelligence.
“Those we know it is the AI on the phone that helped enhance their pictures; we can’t exclude such people because it was the AI on the phone that did that.
“Such people will still be on the register, but those AI-generated photographs, all of these will be removed. We are already looking at them,” he said.
Bayode also disclosed that some individuals allegedly exploited a special registration arrangement designed for persons with disabilities, particularly those with incomplete fingers or no fingers, to introduce questionable voter profiles into the system.
He said the special registration option was among the loopholes allegedly exploited by those responsible for the irregularities.
The commission, he added, was working to eliminate the affected records from the preliminary voters’ register.
Bayode urged Nigerians to rely on INEC’s official communication channels to verify information concerning elections, particularly announcements about changes to election dates or voting arrangements.
“The platform where you can confirm if an election is postponed or cancelled is on INEC media platforms,” he said.
Reacting to the discovery of suspected AI-generated photographs in the voters’ register, a Facebook user, Edos Osayande, questioned the commission’s position on the technology.
He wrote, “But they told Nigerians they are going to use AI for elections na. This is part of the AI.”
Amupitan Defends AI
Meanwhile, INEC Chairman, Prof. Joash Amupitan (SAN), has defended the commission’s adoption of artificial intelligence in election management, insisting that the technology should not be regarded as a threat to democracy.
Speaking at the Beyond the Ballot Summit in Abuja on November 2, Amupitan said AI could strengthen electoral administration through improved voter registration, logistics planning, data analysis and the management of large voter databases.
“AI is not an enemy of democracy,” he said.
However, he acknowledged that the increasing accessibility of generative AI tools had created opportunities for malicious actors to manipulate voters and spread disinformation.
He warned that technologies previously requiring sophisticated resources, including voice cloning and the production of realistic videos impersonating election officials, were now accessible at relatively low costs.
“However, the rapid democratisation of generative AI models has dramatically lowered the barrier to entry for malicious actors; capacities and capabilities that require state-level intelligence resources, such as high voice cloning of public figures and hyper-realistic synthesis of election officials making false announcements, are now accessible at negligible cost. These tools target specific vulnerabilities with distinct timelines,” he said.
Experts Back AI
Speaking on the potential benefits of artificial intelligence in electoral management, a Professor of Computer Science and Artificial Intelligence at Babcock University, Remo, Ogun State, Oludele Awodele, said the technology could help INEC identify irregularities, improve election monitoring and strengthen public confidence in the electoral process.
Awodele said AI-powered biometric identification systems could help identify duplicate registrations and other suspicious entries, while data analysis tools could assist the commission in detecting patterns that might indicate electoral fraud.
He added that AI could also improve communication between INEC and voters by making electoral information accessible in different Nigerian languages.
According to the experts, natural language processing tools could help the commission monitor public conversations, complaints, enquiries and feedback from different parts of the country.
Such systems could also assist in identifying emerging security concerns and election-related misinformation requiring urgent attention.
They noted that AI could support the tracking of electoral materials and improve administrative coordination, particularly in a country with a large electorate and thousands of polling units.
However, Awodele warned that the same technology could be weaponised to produce deepfake videos, fabricate audio recordings, impersonate public figures and circulate false election announcements.
He recommended that INEC adopt watermarking systems and deepfake detection tools while strengthening collaboration with social media companies to identify and remove misleading content.
“There should be a mechanism in place, like some of these detection tools that INEC should be able to use to ensure that they pull down that kind of AI-generated disinformation or misinformation,” he said.
The professor also called for appropriate sanctions against individuals and groups involved in creating, sponsoring and circulating misleading AI-generated materials.
He stressed that effective monitoring, public awareness and accountability would be necessary to prevent artificial intelligence from undermining the credibility of the 2027 general elections.
