Experts have called for greater transparency, stronger data infrastructure and clearer regulatory frameworks as Nigeria explores the role of artificial intelligence (AI) and digital technologies in its electoral process.
They said this at the Open Data Day 2026 themed ‘Data, AI and the future of democracy’ hosted by BudgIT foundation in partnership with Open Knowledge Foundation, Nigeria.
While speaking during a panel discussion, Tope Adenugba, IT support specialist at the Independent National Electoral Commission (INEC), representing Muyiwa Yusuf, the head of ICT, said the commission has invested heavily in technology since 2011 to strengthen voter registration and accreditation.
According to him, the commission first deployed electronic voter registration machines in 2011, followed by the introduction of the Smart Card Reader in 2015 and 2019 to verify voters during accreditation.
However, the commission later developed the Bimodal Voter Accreditation System (BVAS), which uses both facial recognition and fingerprint verification to authenticate voters before they are issued ballot papers.
“Once the BVAS cannot recognise a voter, that means the person cannot vote and will not be issued a ballot paper,” Adenugba said.
He explained that the BVAS device hosts two separate applications which are the voter accreditation system and the results upload interface connected to the INEC Result Viewing Portal (IReV), which allows the public to view election results uploaded from polling units.
Addressing accreditation failures
Despite improvements in accreditation technology, Adenugba acknowledged that the system still faces challenges, particularly with voters who registered many years ago.
“Some voters who registered in 2011 or 2014 now look very different, which sometimes causes the device to fail during accreditation,” he said.
To address this issue, INEC plans to conduct a nationwide revalidation exercise to allow voters update their biometric records and personal information.
“This will give citizens the opportunity to update their data so we can reduce accreditation failures and prevent voters from being disenfranchised,” he added.
INEC is also investing in faster servers to enable more efficient result uploads and has introduced an e-learning portal to improve training for polling officials.
While speaking about concerns around infrastructure challenges such as electricity and connectivity, Adenugba said the commission has developed contingency plans.
Election officials typically move to Registration Area Centres (RACs) a day before voting, where backup power systems are available to charge devices. On election day, officials rely on power banks to ensure equipment continues to function.
INEC also conducts network tests before elections to identify the strongest telecommunications providers in each polling unit area.
“Before every election, our staff visit polling units to test connectivity and determine the strongest network for uploading results,” he said.
AI risks and governance concerns
While digital tools are improving election management, experts warned that Nigeria must build stronger governance structures before deploying artificial intelligence in electoral processes.
Temitope Asama, machine learning researcher at the Machine Learning Collective Africa, said transparency and accountability must be central to any AI system used in elections.
According to her, vendors involved in AI deployment should be independent, unbiased and able to publish verifiable documentation explaining how their systems work, including model accuracy, potential bias and limitations.
“Transparency means that the processes are open to scrutiny, the systems are secure, and those deploying them are accountable,” she said.
Asama also warned that Nigeria currently lacks the foundational infrastructure needed for large-scale AI deployment in elections.
“We do not yet have the concrete foundation required to implement AI effectively in electoral systems,” she said.
Data limitations and algorithmic bias
Frank Eleanya, a tech journalist at TechCabal noted that the success of AI systems depends heavily on the quality and diversity of data used to train them.
He said Nigeria’s data infrastructure remains weak, with many datasets incomplete or biased.
“Even basic demographic realities are poorly documented,” he said, noting that many minority communities are often excluded from data used to develop digital systems.
Threat of misinformation and digital manipulation
Gbenga Sesan, executive director of the Paradigm Initiative, warned that emerging technologies could also be used to manipulate political discourse.
He cautioned that elections could see an increase in coordinated disinformation campaigns, deepfakes and algorithm-driven misinformation.
“Before we talk about using AI in elections, we must clearly define what it will actually do,” Sesan said.
He added that troll farms and coordinated online campaigns have already been used in previous elections to influence public conversation, and these tactics could become more sophisticated with AI.
