<p>Academic cheating in online examinations jeopardizes academic integrity and necessitates prompt intervention to uphold the credibility of digital assessments. Traditional proctoring methods are increasingly inadequate in detecting students’ cheating behaviors, especially in the growing landscape of digital education. This study highlights the pressing need for advanced automated solutions to ensure fairness in online examinations, with Artificial Intelligence Proctoring Systems (AIPS) offering a transformative approach. The research focuses on the application of Machine Learning (ML) and Deep Learning (DL) techniques in developing reliable, Automated Proctoring Systems (APS). This research synthesizes 80 peer-reviewed articles published from 2014–2024 in accordance with predetermined inclusion/exclusion criteria to assess the performance, reliability, and limitations of AI-based proctoring systems (including the hardware). In contrast to earlier reviews, this article provides a more thorough synthesis across multiple dimensions. The findings reveal that advanced ML and DL techniques, such as Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs), better detect cheating by analyzing visual cues, including eye movements, head posture, and facial expressions, as compared to traditional techniques. Key evaluation metrics such as precision, recall, F1-score, specificity, and sensitivity are highlighted where available. The findings advocate for the integration of Internet of Things (IoT) and biometric technologies to augment the security, scalability, and operational efficiency of online proctoring systems. This comparative synthesis informs the development of a unified, robust monitoring framework capable of redefining academic integrity standards in the digital learning era in higher education settings.</p>