Artificial intelligence (AI)-personalized learning is often promoted as a way to match instruction to learner differences, yet its educational value should not be judged only by efficiency or test performance. This narrative review uses algorithmic belonging as a conceptual lens to examine how AI-personalized learning may shape students’ inclusion, identity, and participation in diverse classrooms. Across the reviewed literature, personalized and AI-supported systems are typically associated with adaptation, feedback, recommendation, and learner support, while belonging research emphasizes acceptance, recognition, cultural affirmation, and meaningful participation. Direct studies that explicitly connect AI personalization with belonging remain limited; most available evidence must therefore be synthesized from adjacent work on AI in education, personalized learning, school belonging, culturally relevant pedagogy, and diverse-classroom participation. The literature suggests that AI personalization may support inclusion when it increases access, responsiveness, and student agency, but it may undermine belonging when it stereotypes learners, obscures decision rules, or reproduces cultural and linguistic norms that marginalize some groups. The strongest implication is that personalization should be treated as a social and ethical design problem, not merely a technical one. AI systems are most likely to support belonging when they are explainable, fair, culturally sustaining, and embedded in teacher–student relationships that affirm identity and widen participation. Keywords: artificial intelligence; personalized learning; belonging; inclusion; identity; participation; diverse classrooms; culturally relevant pedagogy; algorithmic bias; educational equity