AI AND DEEPFAKES ON SOCIAL MEDIA: A THEORETICAL ANALYSIS OF THE TECHNOLOGICAL DISRUPTION IN NIGERIA'S 2023 GENERAL ELECTIONS

This study provides a theoretical and qualitative analysis of how artificial intelligence (AI) and deepfake technologies disrupted information integrity, voter perception, and electoral credibility during Nigeria’s 2023 general elections. Conducted between December 2024 and August 2025 in Abuja, Lagos, and other key geopolitical zones, the study draws on secondary qualitative data from online media archives, election reports, and scholarly analyses. It explores how AI-driven content manipulation on platforms such as Facebook, X (formerly Twitter), TikTok, and WhatsApp facilitated the spread of misinformation, political propaganda, and audiovisual deepfakes targeting presidential candidates and electoral institutions. The research employs the Technological Disruption Theory and the Agenda-Setting Theory as its theoretical frameworks to explain how technological innovation simultaneously fosters digital inclusion and undermines democratic communication through algorithmic amplification of false content. Findings reveal that several deepfake videos and AI-generated audios circulated during the 2023 elections most notably, the fabricated voice of a leading presidential candidate allegedly discussing vote rigging, which gained over 1.5 million views in less than 48 hours before being debunked by fact-checkers such as Dubawa and AFP. The study also discusses the problem of false positives instances where authentic political videos were wrongly flagged as deepfakes thereby complicating the verification process and reducing public trust in legitimate campaign communications. Through a qualitative synthesis of verified case studies, the paper establishes that AI-enabled deepfakes posed a multidimensional threat: eroding voter confidence, distorting media narratives, and challenging electoral oversight. However, it also notes that the emergence of AI fact-checking tools, collaborative journalism initiatives, and legal reforms by Nigeria’s National Information Technology Development Agency (NITDA) provide partial mitigation. The study concludes that while AI offers potential for transparency and civic innovation, its misuse through deepfakes underscores an urgent need for stronger digital literacy, institutional readiness, and ethical technology governance. The paper recommends policy interventions that integrate AI governance frameworks, promote cross-platform verification systems, and strengthen media accountability. Ultimately, the research contributes to ongoing discourse on digital democracy in the Global South, positioning Nigeria’s 2023 elections as a critical case study in understanding the socio-political consequences of technological disruption in emerging democracies.