<![CDATA[Purpose of the study: The objective of this study is to investigate the role of AI-powered tutors in assisting students to rectify Newtonian misconceptions and attain a conceptual comprehension of Einsteinian physics concepts, including spacetime curvature, time dilation, and gravity as geometry. Methodology: A conceptual and narrative literature review was performed utilizing databases such as Scopus, Web of Science, ERIC, SpringerLink, and Google Scholar. The utilized tools and frameworks encompass conceptual change theory, constructivism, cognitive load theory, Bayesian Knowledge Tracing, reinforcement learning, virtual simulations, and natural language processing. Main Findings: AI-driven tutors proficiently identify misconceptions, deliver tailored feedback, and present multimodal simulations of relativistic phenomena. They augment conceptual comprehension, diminish cognitive load, elevate student engagement and motivation, and facilitate inquiry-based learning. Recently researches indicates enhanced conceptual precision and acceptance of Einsteinian models when artificial intelligence is incorporated with guided instruction. Novelty/Originality of this study: This study integrates artificial intelligence technologies with conceptual change theory and Einsteinian physics education to propose a systematic pedagogical framework. It enhances understanding by demonstrating how AI operates as a cognitive collaborator, improving conceptual restructuring, metacognition, and accessibility to contemporary physics instead of supplanting educators.]]>
