Technological profiles of primary and secondary school teachers: a data-driven approach to AI training design

Abstract This study analyzes the level of knowledge, use, pedagogical purposes, and ethical concerns related to artificial intelligence (AI) tools among primary and secondary school teachers in Spain during the 2024–2025 academic year. Its main objective is to identify teachers’ technological profiles to design more effective AI training programs tailored to their actual needs. A mixed analytical strategy was applied to the data from a survey of 262 teachers. First, Spearman correlations were used to identify consistent constructs of behavior and attitude related to AI use, concerns, and self-training. Second, K-Means clustering was conducted on 71 variables to detect broader segments of teachers based on multivariate similarity. The results of both analyses were combined into a two-level taxonomy and mapped onto DigCompEdu competence areas. K-Means consistently produced two macro groups : active adopters and cautious/emerging adopters. The Spearman structure refined these into five micro profiles. Advanced integrators, Content and resource Builders, Data and Code, Safety first fundamentals, and Skeptics/Context-Limited Users . The dual approach yields a robust and actionable classification of teachers’ AI readiness. The resulting macro-routes and micro-profiles support the design of targeted, scalable, and competence-aligned AI training for educators.