Nigeria’s AI Influence in Upcoming Elections: Can It Be Mitigated?

Questions over voter register

Nigeria’s Independent National Electoral Commission (INEC) has released a preliminary voter register containing images believed to have been generated with artificial intelligence. Reported warning signs include exaggerated facial features, duplicate names and identities of people who have died since the previous election.

The findings have prompted concern about INEC’s preparedness for the 2027 general elections and the credibility of its procedures. Commission chairman Joash Amupitan said INEC is investigating discrepancies that may have resulted from AI use.

A unit to monitor threats

Amupitan announced a dedicated unit to track AI-related threats, identify deepfakes and help protect the electoral process. He warned that the elections’ credibility could otherwise come under attack.

Nigeria has faced fabricated images, doctored audio and fake videos before. During the run-up to the 2023 elections, deepfakes falsely linking candidates to terrorism-related activities circulated on social media.

Risks to voters and evidence

Lois Ugbede, assistant editor at fact-checking platform DUBAWA, told DW that AI-driven disinformation is serious and becoming commonplace. She said parties were already using inflated crowd images to suggest they had broad support, and warned that such content could affect political debate and voter behavior.

Akintunde Babatunde, executive director of the Abuja-based Center for Journalism, Innovation and Development, said a deeper risk is that people may begin doubting authentic information. He cited a lawmaker who, after making inflammatory remarks at a campaign rally, claimed clips of the remarks had been manipulated.

Trust needs oversight

Mgbeahuruike said voters need confidence that technology is transparent, accountable, independently scrutinized and subject to human and institutional oversight. Voice of Nigeria Director-General Mallam Jibrin Baba Ndace also urged journalists to avoid accepting reports that may later prove false because of AI manipulation.