The use of Artificial Intelligence (AI) in modern education has increased significantly, especially since the Society 5.0 era, which emphasizes the integration of technology with human needs. AI has been used for various learning functions, such as personalized content, adaptive learning, automated assessments, virtual tutors, and real-time analysis of student learning progress. This literature review aims to identify the challenges and opportunities for implementing AI in modern learning across various educational contexts. The analysis of current literature indicates that AI holds significant potential for improving learning effectiveness through the automation of teacher administrative tasks, increased access to digital learning resources, and the development of data-driven personalized learning models. However, key challenges identified include school infrastructure readiness, teacher digital skills, algorithm accuracy, data bias, and ethical issues related to student data privacy and security. These literature findings provide insight that the success of AI implementation in schools is heavily influenced by technological readiness, clear regulations, and increased educator capacity in AI literacy. Therefore, AI implementation is not only a technological transformation but also a pedagogical transformation and digital culture within the education ecosystem. This research provides a theoretical foundation for policy development, implementation models, and further research on AI in future education.