Artificial intelligence (AI) into education is becoming a transformative agent offering new chances for enhancing administrative processes, teaching, and learning. Particularly machine learning (ML) and deep learning (DL), recent advances in artificial intelligence technologies have shown great potential in predicting academic achievement, improving teaching strategies, and so supporting decision-making inside educational institutions. Notwithstanding these advances, there are obvious problems and limits that have to be addressed if we are to fully exploit the potential of artificial intelligence in the field of education. Recent research reveals significant limits like poor contextual adaptability of artificial intelligence models, insufficient integration of emerging technologies like augmented reality (AR), and challenges in improving distance learning. Although the integration of AR into educational systems is still under investigated, current artificial intelligence models usually rely on generalised datasets lacking the diversity of educational environments. The shift to online learning has underscored even more the requirement of solid, contextually relevant models to manage assessment strategies, student interaction, and technology acceptance. By means of a comprehensive examination of the corpus of present literature, this paper evaluates the present position of artificial intelligence applications in education so highlighting research needs and constraints. Emphasising their capacity to solve the discovered challenges, the survey focusses on ML and DL application. By means of analysis of current studies and recommended future research routes, this study aims to offer pragmatic insights and recommendations for enhancing the efficiency of artificial intelligence in educational environments.