2027: Yiaga Africa, others seek safeguards against AI election manipulation

In Nigeria's upcoming 2027 general elections, leading electoral body INEC and organizations including Yiaga Africa have urged stringent measures to counter potential abuse of artificial intelligence (AI). They voiced concerns that deepfakes, disinformation and cyber threats could jeopardize electoral integrity. The call came from a communiqué issued following the 2026 AI and Democracy Forum in Abuja, which convened stakeholders such as electoral bodies, tech firms, security agencies, civil society groups, journalists and policymakers to discuss AI's implications on the elections.

The forum proposed that INEC should release an election-specific AI governance framework, detailing AI applications, associated risks, data sources, protective measures, and the duties of technology vendors. They also urged the commission to devise a transparent protocol for handling election results, encompassing the physical Form EC8A recording, AI-assisted figure extraction, electronic transmission, verification, collation, and discrepancy resolution processes.

The communiqué stressed that human verification and prescribed procedures should remain pivotal to electoral processes, especially in cases where inaccuracies might affect voting rights, result credibility, or public trust. Recommendations included conducting independent pre-election audits, stress-testing result upload and transmission systems, and verifying technology vendors' assertions.

Crucially, AI-generated alerts about duplicate registrations or anomalous results were suggested to prompt human investigation instead of automatic deletion, rejection, or suspension.