<![CDATA[Artificial intelligence has transformed language from a human meaning-making practice into an interaction among computational models, semantic theories, and social values. This systematic literature review examines how scholarship conceptualises language meaning in the AI era and identifies conditions for interpretable, culturally responsive, and ethically accountable machine language understanding. Following PRISMA, the review screened records retrieved from Scopus and sources, synthesising 36 eligible studies through thematic analysis and a Theory–Context–Characteristics–Methodology lens. Evidence shows that scholarship is organised around an interconnected AI–machine learning–natural language processing nexus, with large language models, semantic representation, machine translation, and human–AI interaction as lines of inquiry. Rather than demonstrating that machines simply “understand” language, the studies reveal a shift from formal semantic processing towards contextual, pragmatic, and relational accounts of meaning. Persistent limitations remain: technical performance receives greater attention than linguistic diversity, social consequences, interpretability, algorithmic fairness, privacy, and cultural conditions in which meaning is negotiated. The review also identifies an uneven geography of knowledge production: studies from North America and Europe advance foundational and ethical debates, whereas Asian research foregrounds local applications. Its principal contribution is a multidimensional synthesis connecting computational development with philosophy of language, linguistics, cognition, and ethics, moving beyond narrowly technical reviews of AI language systems. Future research should test culturally adaptive and explainable models across minoritised and multilingual languages, combine computational evaluation with human-centred inquiry, and establish cross-regional collaborations that treat linguistic justice as a core design principle. Meaningful AI requires not textual fluency alone, but accountable participation in context-sensitive human communication.]]>
Copyrights © 2026
