Artificial intelligence in education

a state of the art in the Hispanic American context

Abstract

This study analyzes the state of the art regarding the application of artificial intelligence (AI) in education in the Hispanic American context between 2018 and 2025, through a bibliometric and qualitative review of 1,045 documents indexed in Web of Science. The results reveal a paradigm shift beginning in 2023, driven by the emergence of generative AI (e.g., ChatGPT), which led to an exponential increase in scientific output and citations, accounting for 78% of the total citations in the corpus. The research reveals a marked geographical asymmetry: Spain accounts for 71% of the publications, while most Latin American countries have a marginal presence, posing challenges for intellectual sovereignty and cultural adaptation. Thematically, the field has evolved from technocentric approaches and pilot projects toward ethical and pedagogical reflections, highlighting the validation of models such as TAM and UTAUT, as well as debates on academic integrity, equity, and personalized learning. Key findings include gaps in basic education, rural and indigenous contexts, and risks associated with algorithmic bias and technological dependence. The study concludes that AI should not be seen as a panacea, but rather as a mediating tool to promote a sustainable and inclusive educational transformation, aligned with SDG 4. Future directions are proposed, focusing on hybridization with critical pedagogies, the development of localized AI, and the redefinition of assessment and teacher training.