In recent times, there has been a growing interest in enhancing recommendation systems for e-learning platforms to deliver a personalised learning experience that meets each learner's distinct requirements and preferences. Nevertheless, it is crucial to recognise the ethical considerations surrounding this technology, as it heavily relies on gathering and analysing students' data, which could potentially affect their level of engagement and trust in online learning settings. This study delves into the ethical considerations surrounding e-learning recommender systems and their impact on students' behaviours, particularly in relation to privacy concerns, trust, and engagement within the online educational environment. The research emphasises various influential factors such as the transparency of the platform, previous exposure to cybersecurity education, and students' backgrounds, including their gender and educational level. The investigation involved 602 students from various academic levels across 11 distinct regions in Morocco, encompassing both rural and urban areas. Among the key findings from the Structural Modelling Analysis (SEM) is that the level of transparency on online learning platforms is a significant factor in building students' trust. However, an excess of transparency can have the opposite effect, leading to disengagement as students become overwhelmed with information. The study also revealed an inverse relationship between cybersecurity education and platform transparency, as increased knowledge can lead to a more critical examination of platform regulation and a heightened awareness of privacy risks. In addition, the research found that individuals with higher levels of education have better cognitive abilities and critical thinking skills, enabling them to understand transparency better. Interestingly, those with higher education also tend to have fewer privacy concerns, likely due to their confidence in security measures and understanding of data handling practices.