Nigeria’s Independent National Electoral Commission warned Wednesday that AI-generated disinformation ranks among the most serious threats to the credibility of the 2027 general elections, even as it said funds it had requested for election technology and operations have yet to be fully released.
INEC Chairman Prof. Joash Amupitan, represented by National Commissioner Sani Adam, delivered the warning at the 13th Annual International Conference of the Association of Communication Scholars and Professionals of Nigeria in Abuja, themed “Media, Technology and Elections in Africa: Navigating Digital Disruption and the New Dynamics of Electoral Communication.”
Amupitan said generative AI has made it cheap and increasingly difficult for ordinary citizens to detect fabricated video, audio and images, warning that such content could include a synthetic clip of an election official announcing a result that was never declared, or a cloned voice attributed to a candidate conceding a race still undecided. He said the threat had “migrated from the accreditation machine to the information ecosystem” surrounding it, and that flawless technical performance on election day alone would not protect the integrity of the vote.
On preparations, Amupitan said INEC is carrying out a comprehensive audit of its electoral technology systems, covering redundancy, penetration testing and disaster recovery planning, and is weighing a mock presidential election to stress-test voter accreditation, result transmission and the INEC Result Viewing Portal under conditions similar to election day.
He disclosed that the commission had proposed a substantial budget for technology and election operations, including funds to replace accreditation devices lost or damaged during the last election cycle, but said the money had not yet been fully released. The commission has separately flagged vote-buying, localized electoral violence, hate speech and foreign information manipulation among other threats to the 2027 polls, officials said at an earlier briefing last month.
