The Independent National Electoral Commission (INEC) faces the challenge of detecting digital election manipulation without compromising political neutrality as it prepares to deploy artificial intelligence (AI) ahead of the 2027 general elections.
With more than 103 million registered voters and 176,846 polling units across Nigeria’s 774 local government areas, the commission plans to use AI to detect deepfakes, authenticate political content and identify coordinated disinformation.
However, the effectiveness of the initiative will depend on how quickly INEC can verify suspicious content, correct false claims and demonstrate that its interventions are not politically motivated.
Joash Amupitan, INEC chairman, announced the establishment of the AI unit within the commission’s Information and Communication Technology Directorate at the Artificial Intelligence and Democracy Forum 2026 in Abuja on Wednesday.
“Algorithms will not, and must never, decide the 2027 vote,” Amupitan said, stressing that technology should support electoral institutions rather than replace human judgement.
The unit is expected to use real-time monitoring, deepfake detection software and data analytics to identify emerging digital threats. But distinguishing deliberately fabricated content from legitimate political criticism could prove difficult without clear verification procedures and independent oversight.
An analysis of the initiative by BusinessDay showed that its success will depend not only on the sophistication of the technology but also on INEC’s ability to verify suspicious content, correct false claims promptly and maintain public confidence in its neutrality.
Yunusa Ya’u, executive director of the Kano-based Centre for Information Technology and Development (CITAD), told BusinessDay that the effectiveness of the initiative would depend on whether the technology could address the realities of the country’s digital environment, including the rapid circulation of misleading information across social media platforms and messaging applications.
Ya’u’s assessment highlights the need for an approach that combines technological monitoring with digital literacy, independent fact-checking and public education, particularly among voters who may encounter fabricated political messages without the means to verify them.
Similarly, Aremu Adeniyi, executive director of the Civil Society Organisation for Conflict Resolution in Nigeria (CS-CRIN), said the initiative should be assessed based on its ability to protect electoral credibility, reduce the potential for information-driven political tension and help voters distinguish authentic statements from fabricated content.
The concerns reflect the growing risks associated with false videos, cloned voices, fabricated announcements and misleading claims about voting arrangements, which could influence public opinion, damage candidates’ reputations and undermine confidence in election results before votes are cast.
How INEC’s AI unit could influence the 2027 elections
An examination of the proposed unit’s functions and the challenges identified by stakeholders points to five areas in which AI could influence electoral information management ahead of the 2027 elections.
Experts noted that the INEC’s AI unit, if properly implemented, could shorten the time needed to identify fabricated videos, cloned voices and false announcements attributed to candidates or electoral officials.
The second point raised by experts is that the unit could improve public communication by helping INEC respond more quickly to false claims about polling locations, accreditation procedures, election postponements and alleged irregularities.
Read also: INEC sets up AI unit to detect deepfakes, protect 2027 elections
They added that the unit could also support the authentication of campaign materials and help journalists and fact-checkers investigate coordinated attempts to impersonate candidates or misrepresent their positions.
They further stressed that the AI-assisted analysis could support electoral logistics and digital security.
According to them, these applications may help identify operational risks, but they cannot independently resolve challenges such as inadequate personnel, delayed materials, network failures, or poor coordination.
The stakeholders noted that the initiative could increase scrutiny of political parties’ use of AI-generated campaign materials, especially if disclosure requirements and verification standards are applied consistently to the ruling party and opposition.
Accountability will determine public trust
The experts noted that AI detection systems can make mistakes, especially when processing edited footage, poor-quality recordings, unfamiliar accents and content expressed in Nigeria’s many languages.
According to them, an incorrectly classified video or statement could lead to legitimate political expression being labelled misinformation. If the commission is perceived as protecting particular candidates or suppressing criticism, its intervention could deepen the distrust it is intended to address.
They stressed that INEC will need clear procedures for verifying suspicious material, correcting errors, protecting privacy and allowing affected individuals to challenge disputed classifications.
They added that independent experts and fact-checkers should be able to scrutinise the system’s performance without compromising legitimate security requirements.
The stakeholders stressed that human oversight must remain central to decisions with electoral consequences. AI-generated alerts should trigger verification, not automatically determine that content is false or justify punitive action.
INEC has also called on political parties to adopt an ethical code for AI use in campaigns, media organisations to strengthen digital verification, and security agencies and judicial institutions to improve cybercrime investigations and digital forensics.
These measures indicate that the commission’s approach extends beyond technological detection to the responsibilities of institutions involved in electoral information management.
However, their effectiveness would depend on implementation, coordination and the availability of clear standards for assessing disputed content.
African experiences offer lessons, not guarantees
BusinessDay’s analysis showed that Nigeria is not alone in confronting the growing influence of digital platforms on elections.
South Africa, Ghana, Kenya and Senegal have used combinations of fact-checking, digital monitoring, media verification and civic education to respond to misleading political information.
However, a comparison of these experiences does not establish that all four countries have dedicated AI election units comparable to the one announced by INEC.
Neither does it demonstrate that digital monitoring alone has reduced electoral manipulation or changed election outcomes in those countries.
In neighbouring Ghana, civil society groups and fact-checkers monitored misinformation during the 2024 general election.
Figures attributed to the Ghana Fact-Checking Coalition indicate that it documented more than 100 false publications, with social media accounting for 85.3 percent of the recorded false narratives.
The figures suggest that social media was a major channel for the misinformation documented by the coalition.
South Africa’s 2024 elections similarly highlighted the importance of distinguishing between deepfakes and more conventional forms of misinformation, including fabricated documents, misleading claims and genuine images circulated without their original context.
In Kenya, experiences with online political mobilisation showed the speed at which divisive narratives can spread and the difficulty of responding while protecting freedom of expression.
Senegal’s political environment has also demonstrated the importance of independent media verification and credible public communication during periods of political tension.
Experts noted the common lesson for Nigeria in what happened in other African countries is to understand that technological detection works best when supported by trusted institutions, independent verification, responsive electoral authorities, and an informed public.
Technology alone cannot guarantee electoral integrity
For Nigeria, it was discovered that the challenge extends beyond sophisticated AI-generated videos.
Fabricated election notices, recycled photographs presented as current events, misleading footage, counterfeit result sheets and false quotations attributed to candidates could spread widely even without advanced technology.
The implications become more significant when considered against the size of Nigeria’s electoral system.
BusinessDay’s analysis showed that with more than 103 million registered voters and 176,846 polling units, as disclosed by Amupitan, even misinformation targeting a limited number of communities could create confusion over voting arrangements.
Similarly, fabricated claims about violence or electoral malpractice could heighten tensions if they circulate without verification.
Experts have said that the challenge extends beyond detecting false information to ensuring that verified corrections reach affected voters before misleading narratives gain wider acceptance.
It was also suggested that the proposed Electoral Information Integrity Coalition could help address this challenge by bringing together INEC, civil society organisations, the Nigerian Guild of Editors, the Nigeria Fact-Checking Coalition, media organisations and telecommunications operators.
According to them, such collaboration could enable faster verification and coordinated public corrections. It could also give independent organisations a role in scrutinising information-related threats rather than leaving the commission solely responsible for determining the credibility of disputed electoral information.
This paper, however, observed that the proposed coalition’s effectiveness would depend on the extent of its independence, the responsibilities assigned to participating organisations and the procedures for resolving disagreements over disputed content.
For CITAD and other organisations working on technology and civic participation, the question is whether the initiative can make the digital environment safer without undermining citizens’ rights to question public officials, criticise electoral authorities and express political opinions.
BusinessDay observed that these concerns suggest that technological capacity alone may be insufficient where public confidence in the verification process is weak.
Digital preparations extend beyond AI
INEC’s proposed AI unit forms part of wider efforts to strengthen the technological infrastructure supporting the 2027 elections.
Amupitan disclosed that the commission had procured additional Bimodal Voter Accreditation System (BVAS) units, introduced online voter registration and digitised candidate nominations through its candidates’ nomination portal.
According to him, the portal had processed more than 15,000 nominations across 22 political parties, while the commission had completed 10 of the 14 major operational milestones stipulated by the Electoral Act 2026.
The figures indicated the scale of the commission’s ongoing digital and operational preparations, although they do not independently establish the readiness or effectiveness of the proposed AI unit.
While BVAS and digital nomination systems support specific administrative processes, experts stressed proposed AI unit would operate in an information environment involving political parties, media organisations, civil society groups and voters.
They added that its effectiveness would consequently depend on factors extending beyond INEC’s internal technological capacity.
The concerns raised by stakeholders point to the need for INEC to balance technological innovation with independent verification, public accountability and the protection of legitimate political expression.
With the 2027 general elections approaching, the commission faces a significant test of deploying technology quickly enough to counter digital manipulation while ensuring that the technology itself does not become another source of controversy over electoral fairness.
Related News
