The Independent National Electoral Commission (INEC) plans to deploy artificial intelligence (AI) to strengthen result management in the 2027 general election, using automated systems to prevent incorrect documents from appearing on its result-viewing portal and flag inconsistencies in accreditation figures, it emerged on Thursday.
The commission is also developing a framework for responsible AI use, while technology company Meta says it will work with INEC to promote reliable election information and combat misinformation, including AI-generated deepfakes.
The plans emerged at the unveiling of the AI for Elections Hackathon and Election Development Hub, organised by Rise Networks in Abuja to develop practical technology solutions to Nigeria’s electoral challenges.
INEC’s ICT Director, Dr Lawrence Bayode, said the commission had established an Artificial Intelligence Division within its ICT Department and was developing guidelines for responsible deployment.
He said INEC was exploring AI applications in voter information, electoral data management, operational efficiency, and the detection of misinformation and disinformation.
‘The question before us is no longer whether AI has a role to play in an election, but how we can harness its potential without compromising the integrity, transparency and credibility of the electoral process,’ Bayode said.
He cited the commission’s policy on election technology acquisition and deployment, adopted in September 2025, as part of efforts to guide the use of electoral technology.
At a roundtable moderated by broadcast journalist Seun Okinbaloye, an INEC election technology expert explained that the commission was using AI to address weaknesses identified in result management after the 2023 election.
He said photographs of presiding officers and buildings, rather than polling unit result sheets, had been uploaded to the INEC Result Viewing Portal (IReV).
To prevent a recurrence, INEC trained its AI system on Form EC8A polling unit result sheets so it could identify and reject incorrect documents.
The system also checks whether accreditation figures entered electronically match those recorded by the Bimodal Voter Accreditation System (BVAS) and shown on scanned result sheets.
‘If there are 200 accreditations on the physical result sheet that has been scanned, and you are entering 250, that is inconsistency,’ he said, explaining that the system would flag the discrepancy and block the upload until it was corrected.
He said the technology had been piloted in the Federal Capital Territory, Ekiti and Osun states and was expected to be deployed for the 2027 election.
On extending AI to election security and other sensitive areas, he said INEC would proceed gradually, starting with lower-risk applications before considering more complex uses.
Bayode challenged innovators to develop tools that solve practical electoral problems rather than merely demonstrate technical sophistication.
‘Can we develop AI solutions that do not merely demonstrate technological sophistication, but genuinely solve problems confronting our electoral process?
‘Democracy should never become subservient to technology.’
Okinbaloye also called attention to the need to align electoral laws with technological developments, noting that the declaration of winners remained subject to statutory procedures even as AI could improve result collation and analysis.
He urged lawmakers to understand emerging technologies and consider how electoral legislation could accommodate their appropriate use.
Explaining the scope of the initiative, Rise Networks Chief Executive Officer, Dr Toyosi Akerele-Ogunsiji, said the hackathon and development hub would bring together technology developers, election experts, journalists, public policy specialists and other stakeholders to create locally relevant solutions.
The programme has three components: a nationwide hackathon, a physical hub for developing and testing solutions, and a monthly Community of Practice to sustain collaboration among innovators, electoral institutions and other stakeholders.
The hackathon runs until November 12 and is open to Nigerians aged 18 and above who live in the country.
Participants will use synthetic datasets and technical resources to build solutions with tools including Python, R, MySQL and PHP.
The hub will provide graphics processing units, microcontrollers, Arduinos, potentiometers and ultrasonic sensors for hardware and software development, usability testing and penetration testing without interfering with INEC’s existing infrastructure.
The hackathon’s focus areas include trustworthy electoral information, electoral data analysis, inclusive and multilingual civic technology, cybersecurity and resilience, and election observation and citizen accountability.
Climate resilience and environmental innovation will cut across the themes.
Akerele-Ogunsiji said the initiative was intended to strengthen cooperation between Nigeria’s technology ecosystem and electoral institutions while encouraging solutions that could be safely tested and applied to real electoral needs.
The winner will receive N4 million, while the first and second runners-up will get N2.5 million and N1.5 million, respectively, with other participants receiving various consolation prizes.
Meta’s Head of Public Policy for Anglophone West Africa, Sade Dada, said the company would work with INEC to circulate authoritative election information, including official links to voter information and updates.
She said Meta would use its election operations teams, local partnerships and technology to counter misinformation ahead of the 2027 elections.
Its approach, she explained, was to ‘Remove, Reduce and Inform’: remove content that violates its rules, reduce the spread of harmful falsehoods and direct users to reliable information.
Dada said Meta had invested more than $30 billion in safety and security over the past decade and operated a global election integrity operation involving about 40,000 people. She added that its fact-checking partnerships and digital literacy initiatives would also support efforts to identify misleading information.
Warning that generative AI had made fake audio and manipulated videos easier to produce, she said detection tools could examine watermarks and video metadata.
She urged Nigerians to report suspicious posts rather than share them without verification.
Dada also cited GovGuide Nigeria, a chatbot powered by Meta’s Llama AI model, which provides government information through text and voice in English, Hausa, Igbo and Yoruba.
She said AI-based election tools must remain non-partisan, subject to human oversight and protected by strong privacy and security safeguards.
Developers, she added, should use public, synthetic or properly authorised data when building and testing electoral solutions.
Other participants, including Lagos Business School Dean, Prof. Olayinka David-West, and legal practitioner Tsema Ede, highlighted AI’s potential to improve result management and access to reliable information, while stressing the need for transparency, legal safeguards and public trust ahead of the 2027 general election.