ABSTRACT: 

Despite the numerous benefits offered by Artificial Intelligence (AI) applications, recent developments have also revealed their misuse as a significant enabler of various forms of cybercrime, particularly in the creation of so-called deep fake content. Deep fake technology refers to advanced AI-driven techniques capable of generating highly realistic static images, videos, and audio recordings that closely resemble authentic human appearance and behavior. In light of the rapid and remarkable advancements in artificial intelligence, producing convincingly forged visual and auditory content has become increasingly accessible and sophisticated. This research addresses the challenges posed by artificial intelligence and deep fake technologies, focusing on the algorithms and techniques used to generate manipulated videos and audio recordings that have demonstrated a substantial capacity to influence public opinion, decision-making processes, and public policies. Such fabricated content is often exploited for malicious purposes, including election manipulation, financial fraud, and the defamation of public figures.  The researchers adopted a descriptive methodology to systematically explain the nature and implications of deep fake technologies, alongside an analytical approach to examine existing tools, techniques, and applications used to detect forgery and mitigate its associated risks. Based on this methodological framework, the study presents practical findings and proposes a set of recommendations aimed at addressing deep fake attacks and digital forgery crimes.

Keywords:  Artificial Intelligence Applications, Cybercrimes, Deep Fake, Cybercrime, Information Technology.

ABSTRACT: 

This study presents an analytical approach to evaluating privacy policy compliance in Fintech and Medtech applications using machine-learning techniques. With the increasing reliance on digital platforms for financial and healthcare services, ensuring adherence to privacy standards has become critical. The aim of this work is to assess and quantify compliance levels by examining key data handling features and identifying patterns that influence regulatory adherence. The dataset used in this study was obtained from the profile of Scholars way Research Hub, Uyo, on Kaggle and consisted of 16 relevant features, including Data Collection Practices, Consent Mechanism, Privacy Policy Clarity Score, and Regulatory Violations or Fines. The data was pre-processed and analyzed using Google Colab to ensure accuracy and readiness for modeling. Two machine-learning algorithms, Support Vector Machine (SVM) and Multinomial Naive Bayes (MNB), were applied to classify applications into compliance levels. Model performance was evaluated using metrics such as accuracy, precision, recall, and F1-score. Results showed that the SVM model achieved a higher accuracy of 79%, demonstrating strong capability in identifying low-compliance applications, while the MNB model achieved 68% accuracy with more balanced classification performance. Visualization tools such as heat maps, bar charts, and pie charts revealed significant relationships, particularly between compliance levels and privacy policy clarity, as well as notable regulatory violations and user complaints. Based on these findings, it is recommended that organizations improve privacy policy clarity, strengthen compliance monitoring mechanisms, and adopt machine learning-based tools for continuous assessment of data protection practices.

 Keywords: Fintech, Medtech, NLP, Privacy, Privacy, Compliance

ABSTRACT: 

The rapid growth of smart engineering systems has increased the demand for real-time, low-latency, and intelligent decision-making solutions. Traditional cloud-centric computing models are often limited by high latency, bandwidth constraints, and privacy concerns, making them unsuitable for time-sensitive applications. This paper proposes an intelligent AI-driven edge-computing framework designed to enable real-time optimization in smart engineering environments. The framework integrates edge intelligence, adaptive task offloading, and artificial intelligence techniques, including reinforcement learning and federated learning, to enhance system performance. A multi-layered architecture comprising sensing, edge intelligence, edge–cloud coordination, and application layers is developed to support efficient data processing and decision-making. The system model is formulated to minimize latency and energy consumption while maximizing throughput and meeting Quality of Service (QoS) requirements. Simulation results demonstrate that the proposed approach significantly outperforms traditional local, edge-only, and cloud-based methods, achieving reduced latency, lower energy consumption, higher throughput, improved deadline satisfaction, and more consistent performance. The framework is applicable to various domains, including smart manufacturing, intelligent transportation, smart grids, and healthcare systems, providing a scalable and efficient solution for next-generation real-time applications.

Keywords:  AI-driven edge computing; Edge intelligence; Real-time optimization; Smart engineering systems; Task offloading; Reinforcement learning; Federated learning; Internet of Things (IoT).

ABSTRACT: 

The increasing integration of renewable energy into industrial power systems in Africa presents significant challenges in maintaining reliability and resilience, particularly in weak-grid environments with frequent disturbances. This paper proposes an AI-enabled predictive control framework for enhancing the resilience of low-carbon industrial energy supply systems. The approach combines machine-learning-based short-term forecasting of renewable generation and load demand with model predictive control (MPC) to optimally coordinate battery energy storage and grid interaction. A unified MATLAB implementation is developed, enabling reproducible evaluation under multiple operational scenarios, including baseline operation, grid outages, and industrial load surges. Simulation results demonstrate that the proposed framework maintains a high resilience index (>0.9), effectively mitigates unserved energy, and maximizes renewable utilization, even under severe disturbances. The results confirm that AI-assisted predictive control provides a practical, scalable solution for improving reliability, sustainability, and operational flexibility of industrial energy systems in regions with grid limitations.

 Keywords:  Resilience, Low-carbon, Grid outage, load surge, BESS, Artificial Intelligence.