Spectra Human-Centered Learning Framework: A Socio-Technical Architecture for AI-Supported Higher Education

Abstract

Generative artificial intelligence (GenAI) can produce persuasive university work without revealing who performed or checked the reasoning behind it. This conceptual article develops an AI-specific elaboration of the previously disseminated Spectra Human-Centered Learning Framework. Its three foundational conditions, six pedagogical phases and five original propositions are retained, not presented as new inventions. A purposive, critical synthesis distinguishes the functions of the pedagogical cycle and examines how AI assistance, intellectual responsibility, institutional safeguards and contextual constraints may intersect with them. The resulting architecture connects real-world input, conceptual framing, engagement, production, assessment and reflective transfer. The inherited propositions remain verbatim; task-level interpretations, candidate indicators and conditions that might challenge them are added for future research. Augmentation, dependency and substitution describe provisional distributions of work between learners and systems, not validated learner types. A targeted comparison with adjacent frameworks identifies overlaps and the specific scope of this elaboration without claiming global novelty or superiority. No student data were collected, no intervention was implemented and no systematic literature search was conducted. Whether the configuration supports independently defensible learning, accessible assessment and transfer across contexts remains an empirical question.