Abstract: The integration of machine learning–based personalized learning systems in secondary education offers significant potential to enhance instructional differentiation, learner monitoring, and academic support. However, adoption depends not only on technological capability but also on institutional readiness across infrastructure, human capacity, and leadership dimensions. This study evaluated institutional readiness and barriers to implementing ML-based personalized learning in secondary schools in Rwanda using a mixed-methods design. Data were collected from 200 institutional staff members across ten secondary schools representing all five provinces. Findings indicate moderate but uneven readiness. Infrastructure indicators show that 55% of schools reported functional ICT laboratories, yet only 42% reported reliable internet connectivity and 36% adequate technical maintenance support. Human capacity gaps were pronounced: while 84% of respondents rated staff digital literacy as high or moderate, only 18% reported high understanding of ML concepts and 22% expressed high confidence in using AI tools. Governance readiness was limited, with 61% reporting data protection policies but only 29% indicating the presence of an institutional AI or digital strategy, and 63% lacking regular data quality audits. Leadership openness to innovation was relatively strong (52%), yet only 33% reported strong ICT budget allocation. The most significant barriers identified were limited technical expertise (64%), insufficient funding (58%), and data privacy concerns (52%), whereas resistance to change was lower (29%). The study concludes that ML-based personalization adoption is constrained primarily by structural capacity and governance gaps rather than institutional resistance, highlighting the need for coordinated readiness strengthening across schools.

Abstract: In an era where traditional cybersecurity measures are increasingly inadequate against the growing sophistication and volume of digital threats, banks are compelled to adopt innovative technologies such as AI. The study explored the adoption and effectiveness of Artificial Intelligence (AI)-based strategies for data protection among banking institutions in Uasin-Gishu County, Kenya. The present study employed surveys and interviews to investigate the integration of AI tools, particularly machine learning algorithms for fraud detection, predictive analytics, and anomaly detection in selected banking institutions. The study revealed that all surveyed banks (100%) had adopted AI-based security systems, with 68.6% focusing on fraud prevention and 66.7% on anomaly detection. A statistically significant relationship (p < 0.001) was observed between strong data protection policies and firms’ integration of AI in security frameworks. The adoption of AI was driven by its capacity to predict threats, enhance fraud detection, and improve operational efficiency. Despite these benefits, the study identified significant challenges, including a shortage of skilled professionals for AI system implementation and persistent concerns regarding data privacy. Additionally, existing regulatory frameworks were deemed insufficient to address emerging risks associated with AI-driven data security. Most respondents acknowledged that the advantages of AI outweighed its challenges, making it a preferred solution for enhancing data protection. The study concludes that AI enhances data security in the banking sector and recommends strengthened regulatory frameworks, increased investment in specialized AI training, and continuous stakeholder engagement to maximize AI’s potential in the cybersecurity landscape.

Henrik Hansson – Department of Computer & Systems Sciences, Stockholm University, Sweden.
henrik.hansson@dsv.su.se

Abstract: The increasing prevalence of deepfake videos poses significant threats to information integrity, political stability, and public trust. This study presents a dual-faceted approach: (1) developing a machine learning model for detecting deepfake videos using visual features extracted from benchmark datasets, and (2) conducting a trend analysis of deepfake content dissemination on social media platforms such as YouTube and Twitter (now known as X). Conducted using the FaceForensics++ dataset and metadata from over 2,000 social media posts collected between 2018 and 2024, this study used a fine-tuned Xception model and natural language techniques. Key findings indicate a post-2020 surge in politically motivated deepfakes and platform-specific propagation patterns. It is recommended that stakeholders implement real-time detection and awareness tools to mitigate social impact.

Abstract: The study focused on assessing the impacts of Artificial Intelligence (AI) through Machine Learning (ML) and Deep Learning (DL) modes in combating the COVID-19 pandemic. This study performed a systematic literature review by analyzing articles published between 2020 and 2023 using VOSviewer (version 1.6.18), MS Excel, SPSS, and the PRISMA 2020 statement. The result shows that the research began in 2020 with 105(8%) articles and increased to 480(35%) articles by 2021, in year 2022 a remarkable increase occurred to 622 (46%) publications. The publication trend declined from 46% in the year 2022 to 11% in the year 2023, this might be due to the decrease of pandemic infections. The study includes only the review of the research articles published from 2020 to 2023 and indexed in the Scopus database. The study has explored the impacts of ML and DL models in combating COVID-19 pandemic; that encompasses a rapid and adaptive response to the urgent needs of the pandemic, the discovery of disease and treatment. The study helps society to improve efficiency, enhanced convenience, and innovative solutions across various sectors, such as the ability to react to health emergencies, examine the healthcare and disease management methods. The study provides a remarkable contribution of AI in addressing the Covid-19 pandemic in work a work place and it has been used to monitor and manage the pandemic on a global scale. In addition, the study contributes to understanding the recent growth statistics of global publications that address COVID-19 pandemic.

Abstract: The main objective of this paper is to assess the possible uses of artificial intelligence (AI) tools and applications for information professionals employed by Tanzania’s higher education institutions (HEIs), particularly in the areas of end-user services and information technology where AI tools and applications may be employed. To find out what they thought about the use and adoption of AI techniques by information professionals of HEIs in Tanzanian, ten information officers were surveyed using a qualitative method. The results’ descriptive tabulation demonstrates that information professionals are knowledgeable about AI tools and applications. Information professionals use Natural Language Processing (NLP)-based applications like Google Assistant, Voice Search, and Google Translate. Pattern recognition methods, such as text data mining, are also used for online searching and information retrieval. OneDrive, Google Drive, cloud computing, and other services can all provide access to Big Data (BD). The population has very little knowledge of chatbots and robotics. The study guides how to use AI technologies and tools for information professionals who have not yet embraced or deployed them or wish to employ more advanced solutions. To establish AI laboratories for information professionals and information science in general, Tanzanian HEIs should collaborate with departments, computing units, and subject matter specialists. Lack of funding and technological know-how is one the biggest challenges facing information professionals in HEIs when implementing AI technologies and solutions.