Journal of Advanced Robotics, Autonomous Systems and Human-Machine Interaction
Artificial Intelligence and Electoral Cybersecurity: A Technical Case Study of South Africaâs 2024 National and Provincial Elections
Abstract
John Maphephe and Surendra Thakur
Artificial Intelligence (AI) is becoming increasingly central to electoral cybersecurity, offering advanced capabilities for real-time anomaly detection, incident response, and disinformation monitoring. This paper presents a technical case study of the Electoral Commission of South Africa’s deployment of AI during the 2024 National and Provincial Elections (NPE). The Commission implemented AI-enhanced firewalls, intrusion prevention systems, endpoint detection and response (EDR), web application firewalls (WAF), and a parallel Security Operations Centre (SOC) to detect, analyze, and mitigate cyber threats. Advanced tools—including User and Entity Behavior Analytics (UEBA), phishing detection, dark-web monitoring, and disinformation tracking—were operationalized at scale using AI and machine learning (ML). Natural language processing (NLP) further augmented threat anticipation and intelligence capabilities. The analysis demonstrates that AI served as a critical force multiplier in safeguarding electoral infrastructure, while highlighting ongoing challenges related to transparency, dependence on external providers, and the explainability of AI-generated outputs. This study contributes to the technical literature on AI-driven cybersecurity by detailing architectural integration, evaluating operational effectiveness, and providing evidence-based recommendations for enhancing resilience in future elections.