Dataset for: Human-Centred Artificial Intelligence in Rural Education: Situated Knowledge, Community Responsibility and AI-Resilient Skills in Villarroya de los Pinares

The rapid expansion of generative artificial intelligence has intensified a central question for education: what should remain human, situated and socially protected when automated systems become available for everyday problem-solving. This question has particular significance in rural communities, where learning, care, work and local decision-making are sustained through proximity, intergenerational knowledge, trust and responsibility for place. This article presents an instrumental qualitative case study in Villarroya de los Pinares, a rural municipality in Teruel, Spain. The study explores which human competencies, dispositions and forms of situated rural knowledge emerge in participants' discourses as difficult to replace by artificial intelligence, and how these meanings may refine the AI-Resilient Skills model from a rural-community education perspective. Data were generated through four semi-structured interviews with community actors connected to local governance and livestock farming, rural entrepreneurship and tourism, cultural mediation and primary education. The analysis followed a thematic-discursive approach through units of meaning, combining inductive coding with a deductive dialogue with the six dimensions of the model. Findings suggest that what remains human is not only cognitive performance, but situated judgement, relational care, local memory, embodied presence, intergenerational learning, critical collaboration with technology and ethical responsibility for decisions affecting people and territory. The article contributes to rural education scholarship by reframing artificial intelligence as a contemporary tension through which the educational value of rural knowledge, community engagement and human-centred decision-making becomes visible.